Ask HN: What's a promising area to work on?

1331 points by richtapestry ↗ HN
What's an area that people think is up and coming? (e.g. like social networks were in 2004, mobile apps in 2010, or vlogging in 2014)

I've finished several projects simultaneously and I'm looking to work in an area with lots of users, but as yet few producers. Wouldn't even need to involve programming, but probably would need to be online, as I'm pretty introverted!

946 comments

[ 2.7 ms ] story [ 329 ms ] thread
Machine Learning is the hot thing right now. Just find an area you like, and apply it. Not sure what is after this.
Thanks, I've had it on my list of stuff to look into for a while. Any advice on where to start?
A well honed skill set in 'ML' isn't even about the algorithms or data structures or any of that. If you want to be competitive in ML, study and think about experimental design, and what it takes to be able to confidently state the accuracy, precision, and reproducibility of a given analysis when you apply it to real world data.

The biggest issue/ liability with most 'AI'/'ML' is unsubstantiated/ irreproducible results that come from not being rigorous with ones experimental design.

I'm a complete beginner in this area, this sounds like a great aspect to bear in mind when learning, thanks!
How do you go about developing skill in experimental design?
Your local university stats department may have a course on it, or check out a textbook. There may well be some online courses too.

Unfortunately I wasn't very fond of the textbooks I used in school and I haven't looked at any since so I can't give you any recommendations.

Probably Kaggle competitions
Read "Data Analysis: A Bayesian Tutorial," by D.S. Sivia for an excellent introduction to experimental design for scientists. This gives you a quantitative approach to the design of experiments, with tools to properly evaluate your results.
I would also recommend the bayesian statistics book with the dogs on the cover. Can't remember its name off hand, but got handed a copy. Its a great foundation.
I think the first step is to develop a skeptics mindset. As a scientist, your job isn't to believe, its to address evidence and evaluate whether or to it supports or conflicts with a given hypothesis. A profound influence on my development in the sciences was a older (he was 55+ at the time), physical chemist I shared an office with,'John'. Arguably, he was one of the greatest critical thinkers I had ever had the privilege of working with. Many other people at that office would avoid him entirely when discussing their work because if John heard about what they were doing, he would always 'challenge' (politely and professionally) them on what they were doing: how they had decided to set up their experiment; how they had decided to sample; how they intended to analyse their data; how they (physically and literally) planned on getting their data.

John did this because on his own work, he was constantly challenging his own assumptions. The result of this was that John 'appeared' to be far less productive; however, the results of his work were orders of magnitude more robust than our other colleagues at that time.

By incorporating a skeptics mindset, less becomes much more. A skeptic doesn't believe the internal error rates produced as a result of a model run, they take a long time to become convinced that a 'thing' is the truth. Don't believe, measure.

How does this apply to ML, beyond the usual train/validation/test split?
Because, the usual train/ validation/ test design often fails to generate a useful ML model/ pipeline. With out some serious consideration into the nitty gritty of the 'experiment design' (see above for what I mean by that), we get 'all-hat no cattle' results.

Lets take a small toy example from one that came up a few days ago, the model that could predict 'heart-disease' from one heartbeat. Their data set came from two different sources: their 'disease state' and 'null state' patients had their cardiograms recorded via different instruments. They did some statistical re-sampling to get the data to 'match', so, a skeptics flags should already be raised. Second major issue: without resampling, they had an effective N of 30; they made the decision to 'slice' every ones cardio gram into thousands of examples. The RNN (i think it was an RNN), obviously, will need thousands of examples to train. They then randomly sampled (according to their train/validation/test split) from this distribution of 'beats' to train their model.

So they didn't do anything 'wrong' according to what myself, and what I would assume yourself were told when we did this or that course in ML. But actually, from an experimental design POV, this is clearly going to overfit. Even disregarding the resampling of the original data (its own, very suspect issue), the slicing alone is enough to realize 'Ah. This is horse crap'. Think about it like this. Say its a 80:10:10 split. Take a random sample of n heartbeats, 3 heartbeats long, from n=100k heartbeats. What is the probability of a heartbeat in the training data set not being very close (in time proximity) to a heartbeat in the validation/ test set? I don't have the time to work it out on paper, but its most likely that anything in the test/ validation set will be temporally adjacent to something in the training set. The probability of anything in the test/ validation being very different from something in the training dataset (say, 3-4 beat sets away in either direction) is very very unlikely. The vast majority of data in the test/validation dataset will have two direct neighbors, both used in training, the next most populated class will have 1 neighbor in one direction, and the almost none will be separated at great distance (even 2-3 beats isolated from something in the training dataset).

This issue central here is the lack of independence in the experimental design. They've created quasi-independence in their sampling methodology, but at the end of the day, they've still only got an n of 30 (not 300,000-3,000,0000). I get it. For most cases, one almost always has to create a condition of quasi-independence in ones data. To get these algorithms to work, you need lots of data.

Knowledge of what ML is, how it works under the hood (a bit), and how to implement it (more imp. imo), all matter in this space. What matters more (imo), is the mindset that can be developed by doing science as an intellectual exercise. Its good science to remain skeptical and be adherent to the evidence rather than our assumptions.

Machine Learning for Humans.pdf
fast.ai has some great intro courses on deep learning.
Assuming this is sarcastic. If so, A+ would read again.
Machine learning was up and coming in 2012 when I graduated college. Not to say it is a bad career choice, just that it's safe to say that ML's time has already come.
(comment deleted)
What are some examples of new areas to apply machine learning?
Any context where automation has been held back by weak perceptual capabilities.
But will machine learning experts be automated out of a job by their own algorithms?
Probably image processing.
Real time raytracing
Can you elaborate? I’m not familiar with the current state of graphics research. Can a high end GPU today do this, and if not, how far are the fastest high-quality algorithms from realtime today?
A high end GPU today (like GTX 10 series) can absolutely stun! Check this out,

http://www.realtimerendering.com/raytracinggems/unofficial_R... [33MB PDF]

It sheds light on some ray tracing techniques with the focus on the new RTX API with DirectX12

Real time rendering portal page generally my goto for checking out new stuff in graphics and now they have a Raytracing section https://www.realtimerendering.com/raytracing.html

I am on phone right now, and will try to put more material as soon as I get back to my workstation!

Edit: Courtesy Onion2k

Note to users on mobile - link is a 33mb pdf.
Nvidia has a popular series of "RTX" gpus released in the last year, which are billed as having real-time ray-tracing capability of some sort. It is not exactly clear to me what the best ways of making use of this are yet; if you want to path trace a whole scene it's still absurdly expensive / low frame rate (check out Quake II RTX) but there are a small number of AAA games released in the past year which are making use of real-time raytracing in small amounts for more subtle graphical improvements.
Given slowing growth of computing power, and that ray tracing algorithms are already mature and well studied, I don't think high-sample real-time raytracing will come along for a long time.

Instead, I think de-noising based on machine-learning will be the crucial factor for real-time raytracing, as Nvidia has been showing.

I think a coming advancement will be in dispatching rays based on ML too. That is, economically deciding where to cast rays, based on feedback from the de-noiser.

Gaming on the cloud. VR and AR.
This is a great one, AAA game install sizes have been ballooning while storage space has plateaued and network speeds and connectivity are getting better.

Maybe develop some split model, where assets are streamed over the network but gameplay and rendering are local?

Or do most things on the server, but stream a spherical color and depth map, so the client can handle rotating the camera with zero lag, and then you can perform some compensation if the camera translates (similar to how Oculus does motion compensation)?

Still white elephants after decades. Every few years someone tries to stream games, it doesn't work.

The closest example right this moment is Microsoft's flight sim that'll download world data from bing maps as you fly around. Simply because they are using synthesized petabytes of map data. Guildwars and other mmo's download world data as you are playing to fill out areas you will visit in the future. It still uses a traditional game client and business model. There's a slew of games that run on nearly every bit of hardware. I don't see the advantage cloud gaming brings.

You could stream every mobile game that's existed as low latency is rarely a requirement for those games. Those games are still downloaded to memory and rendered on the client and use a conventional business model.

Vr is getting a valve style game soon that might well be a killer app for it, but the problem of movement remains.

Real-time communications.

There's a hell of a lot more to building and scaling voice and video than you think, and tons more parts of the world are going to "come more online" with higher-bandwidth telecommunications access in the coming decades.

As to your question of promising areas, I see two major opportunities in health that are coming online in the near future:

* psychedelics as medicine

* gut/microbiome

seconding both of these.

psychedelics as medicine are finally starting to gain the traction and serious research which they will need to be utilized therapeutically. while imperfect, they're potentially curative therapies which are likely a quantum leap ahead of much of what psychopharmacology has had in its roster so far.

the gut/microbiome stuff is also starting to come to a head after about 9 years of intense research activity. clinical applications are right around the corner, and there's an entire universe to explore which we've only scratched the surface of so far.

i'd also like to add cannabinoid (CBD) therapeutics as an addendum to psychedelic medicines. while there's a tremendous amount of hype and folk science surrounding CBD, it appears as though the chemical does in fact have a massive impact in domains ranging from immunology to psychiatry to geriatrics. we're just at the very start of the research process, and there are already a cornucopia of compelling and empirically valid leads which need to be followed.

Let's not forget: - non invasive medical imaging, particularly in terms of new approaches to generating or interpreting CT/MRI/ultrasound scans - more ergonomic surgical robots that don't hold their instruments like a 3 year old toddler - AI/expert systems for rapid triage and assessment of patients - more effective ways of drug delivery (both cell based therapy and developing better oral formulations)
Psychedlics will be useful but hard to monetize, but lol @ microbiome.

You realize what this is, right? High throughput sequencing + multiplexing makes it cheap as chips to start doing -omics on anything you can write a grant for. So everyone and their mother starts getting funding for doing the most boring possible experiments that are total fishing expeditions. The results: we see ... stuff! Isn't that great!? We've got correlations coming out of every conceivable orifice, must be something interesting there, right!?

It's great for writing shitty headlines, but piss-poor for doing any sort of science. Seriously, try talking to any 'microbiome' expert about microbiology. It's scary!

Here's what will happen; we'll get some minor insights into diet wrt health, a handful of treatments for some specific niche gut issues (e.g. fecal transplant), and some handwavy BS about inflammation that will only be meaningful with more work on good old-fashioned signalling from cell/mol. biology.

Of all the scam-y biotech startups (which is basically redundant), microbiome stuff is by far the worst.

Agreed, but there is lots of competition on the second one so if I were starting from scratch I would work on the former.
You need to take a longer term view if you want to notice things before they become trends. Hopping on the bandwagon once it's a "hot" or "promising" area is too late.
Security and privacy. Demand for security folks is white-hot right now, and the need for better security products continues to grow.

I've been seeing a lot of companies in adjacent spaces making headway into the security domain as well—ML assisted intrusion detection and cloud-centric security tooling come to mind.

In a broader sense, I think more governments and regulatory bodies are starting to take cybersecurity much more seriously than they did in the past. GDPR and CCPA are likely the tip of the iceburg in coming year and—regulations aside—I think there is a real and pressing need for better privacy and protection in both consumer and enterprise contexts.

+1 on this; I also see an upcoming trend with tech like deepfakes that make it hard for individuals to tell what is real and what isn't. This intersects with personal privacy when anyone can grab your FB profile pic and then make you do anything in a deepfaked video and destroy your life.
This is the most "for sure" answer in this thread. If I had to bet on any one of these answers being true, this is the answer I would bet on.

Security has always been a big deal, but in light of all the attacks we're seeing lately, there's likely going to be REGULATIONS involved, that may mean demand for these jobs skyrockets.

I have a hunch that the Next Big Thing (TM) will be augmented reality: though the technology is currently in its infancy, when compelling AR glasses are finally released (Apple is rumored to be released theirs in 2020) I think that it will have a massive impact.

AR frameworks are a dime a dozen, but if you're an Apple person look into ARKit + Swift.

I agree -- on top of this, I think machine learning will play an important role. The power of machine learning + AR creates some interesting possibilities. You can check out our website for some examples https://www.2020cv.com
What value does augmented reality add to anything?
I'm thinking of assisting drivers. Projecting stuff into the road to guide the driver, or warn him when there are dangerous objects.
I'm not sure the thing people have been doing perfectly well for 100 years without AR glasses that billions of dollars are being invested in making people not need to do at all is the place to launch your AR dream company
For values of "Doing perfectly well" that encompass "Killing 30,000 people a year."
I like the idea of using it for safety. Is that something currently being done, or still in the experimental stages?
If you don't need a monitor, you don't need the computer as we know it today. Combine it with a brain computer interface and you throw away the keyboard and mouse as well. Subsequently you also throw away all the OS'es as we know them. AR/VR is a toy today. In the future our computers and mobile phones might look as arcane as an eniac.
I recommend Hegel with Neuralink: https://youtube.com/watch?v=ppGlEiQXrRI

Unless the computer is itself capable of anticipating and satisfying your desires — which, I grant, is possible, but, I would argue, incredibly, incredibly dystopian — you will need to consciously articulate your thoughts in such a way that the computer serves your will. That conscious articulation is the bottleneck. Not the keyboard, not the mouse, and certainly not the typing speed.

This is, of course, assuming that the computer remains the servant rather than becoming the master.

If it needs to be a learnable skill like typing, that is just as revolutionary. The interface does not have to be that deep. Just enough to input words and movements. Hell it could be voice commands.
Not if the BCI is trained on data produced by users in a natural thinking state.
I am thinking of the keyboardless typing with wrist bands as an intermediary step. I agree controlling conscious thought is difficult, but you could just key in commands with small impulses of the wrists and it would still be quite a bit more convenient than pulling out the phone
Adtech. Hence Facebook/Oculus.
Think about car mechanics seeing what do next if you want to open this specific motor or electrician seeing where cables are in the wall. Or physicians seeing specific information about patients eg how fast is he is breathing, what patient temperature is etc.
Ask the porn industry in 1-2 years.
Visualizing any change to the exterior of a building or new construction project in the actual location that it will be.
And how many people would need that on a daily basis? This sounds to be a useful but somewhat niche area
There are 1.2 millions houses built in the US every year. I could see the applications pretty vast actually. Another possible idea is for landscaping. Often times a tree will be planted that ends up too big or too small for the space. With an app that can show a tree in the space at various stages in it's life, it could really help in landscape design.
One of the big wildcards for widespread consumer use is the social aspect. We can envision what AR might look like in everyday life. What's far less clear to me is the degree to which people will accept AR glasses that are constantly using video and audio to deliver information to the wearer will be accepted. If I know that person I just met has inconspicuously scanned me and looked up all kinds of information that are now being displayed, am I OK with that?
What if last week that person you were introduced to, and gave your name to(or didn't) happened to look you up and do some research. It wasn't handed to them through glasses immediately, but essentially it's the same thing. I would bet it probably has already happened to many of us and we didn't even realize.

ever have someone come up to you and say "Oh your the person who x"

just a thought on your last sentence

A little tangential, but I hope the answer to this is a resounding "no", and that adoption is sufficient to make a lot of people really uncomfortable. Perhaps this is naive of me, but I still believe that a great contributor to tolerance of our society for massive dragnet surveillance is that it isn't very visible.
Not tangential at all. On the one hand, I can see AR that's extremely powerful and useful given wearable hardware, sophisticated software, image recognition, connectivity, etc.

On the other hand, that would clearly be a step beyond today's smartphones to, as you suggest, potentially always on video and audio that's constantly communicating with and being analyzed by databases.

By contrast, VR seems to me to be a niche.

I think AR almost has to be Apple, not from a tech perspective, but from "this tech is very creepy, invasive, and visibly so" angle. Apple may be the only company that is simultaneously big enough to make the tech happen and charismatic enough to get us to let down our guard to it. Although that latter one is rapidly diminishing.
That's an interesting angle. Though I'm not sure to what degree the average consumer draws this distinction between Apple and Google.

Apple also probably just has about as much brand permission as anyone to create a new category of consumer device and shepherd it through the first couple of versions that will doubtless have shortcomings.

When I say charisma I mean literally being charming, like Google didn't have the tact to role out glass, because glass _looked_ creepy. Apple would have had the tact to know not to. Furthermore I think we're looking at the first couple of pseudo-generations of of Apple AR glasses in the form of the latest iPhones and Apple Watches. Watches are testing the hardware that will be colocated with the UI and the A13 the overpowered chip responsible for doing the more powerful CV functions in like the users pocket.
Yes, I suspect we'll see AR on phones/watches well before we'll see it in glasses and associated wearables. There are some AR-ish apps today but they're very limited.

There are a lot of challenges to get glasses right--both from a hardware and a usability/acceptability perspective. But people are already used to using their phones for things so it seems a very natural transition.

Well it would be socially acceptable to be wearing one while working on-the-job for example, a construction worker, doctor, or even a motorcyclist.
These "privacy vs features" questions of new technologies keep coming up. Over and over, people seem to shrug their shoulders over the privacy violations while they eat up the new features.
My guess on this is that people will pretty readily accept it if it provides something valuable to them, something they can't really get without it. The amount of privacy that people have given up using internet tech is what leads me to believe that this will be no different. I remember when people were afraid to use their own names nearly anywhere. Then they began to post their real identities in facebook (and elsewhere) feeling that they could choose who viewed the information. Now people post their real identities, pictures of themselves and even their children on publicly viewable instagram accounts. This sticks out to me because I remember thinking in the past that there was some barrier between what people posted publicly vs privately in particular with regards to their children. Not everyone is so open but it is not something I hear discussed amongst my non-programmer friends and a large number of them share in this behavior.
Agreed. With 7nm fresh off the presses and 5nm on the way we're should start seeing some efficient SoC's to do the 3D mapping AR needs. Even if glasses aren't ready to be the primary interface - phones with the hardware to accurately map and overlay the world still has a lot of promise.
Big agree. Smartphones have more or less flatlined in the last couple years and the only viable next step seems to be AR. Snap Inc and a lot of smaller players in the market know this, and their pivot to AR is arguably what has caused their resurgence in the past few months.
I run a VR startup so I'm biased, but: AR and VR. Different side of the same coin. But first (next 5-10 years) VR will be more mainstream, because it's easier to solve from a technological point of view.
In what sense are they different sides of the same coin? I see AR as very practical and VR not so much. I look at everyone walking down the street trying to simultaneous look down at their phone and do other things (walk, drive, etc) and see a huge TAM for AR. Meanwhile, in 1993 I went to the local fairground in my small town and experienced VR-enabled multiplayer game of Doom. It had much lower fidelity than VR today, but if the market really wanted it, there would have been a constant pull for their technology to improve. Instead it's just been fits and starts for the past 26 years since I experienced that demo. Big tech has pumped huge money into VR the last 5 years and the uptake is just not happening.
> In what sense are they different sides of the same coin?

One should look at AR and VR as a spectrum. Eventually, you'll be able to mix and match with the same headset.

> Instead it's just been fits and starts for the past 26 years since I experienced that demo. Big tech has pumped huge money into VR the last 5 years and the uptake is just not happening.

This is incorrect. The Oculus Quest is super successful measured on any conceivable metric, outperforming (sales and retention-wise) any other headset that came before it.

One key barrier to adoption of VR (or any consumer tech) is convenience. The Quest is the first headset that makes it convenient to use VR, as you don't need a PC, no external trackers, but you still get the same level of immersiveness.

There are other areas that need improvement until it speak to a bigger TAM (such as improved resolution, more software, better performance, better ergonomics) - but those are all problems that have a more or less clear path of solution.

In 10 years, everyone will be working in VR headsets, not laptops.

Event sourcing. See dataintensive.net

TL;DR - rip the writeahead log out of databases and materialize not just SQL commands but events yourself, returning control to reality rather than systems.

I think in the same lines (I'm building a relational lang http://tablam.org) and think a modern RDBMS could be made alike:

    command  "city.define miami" |> event pre-process (ie: validations and transformations!) |> write-ahead log |> fire-observers: [post log for later OR block writes]
The idea is to put some of the stored procedures in the "event pre-process" before commit to write-ahead. In the "fire-observers" steps is possible to directly commit to a table (sync) and get immediate feedback (ie: primary key duplicate) OR store the log to be processed later.

The idea is that is possible to send "events" that are not necessarily "insert into table" or similar, just pure business stuff. BUT still possible to operate as always!

Not sure if this is what you mean but these exist - Oracle Queues, MQSeries, RabbitMQ (open source MQSeries copy), all part of a generic thing called Enterprise Service Bus
If the queue is outside the RDBMS, not.

What I say is that RDBMS are too much of a blackbox. I want a way to use them alike legos (but without gettin INTO their source code!).

Imagine you wanna do a bloom index, but the RDBMS not have it. How you do it normally?

Put a handler/listener/event etc that intercept the calls and route if need:

    search ... | if_bloom index_bloom OR continue
I would suggest thinking more along the lines of Kafka/Kinesis/et cetera. ESBs are often not persistent and as a result not available for reprocessing against new use cases or with fixed code.
Govtech and defense tech

Super high barriers to entry, governments are spending more and more but people are demanding better more efficient services and won't put up with waiting in line at the DMV forever. In defense, it's hard to see the applications and what to build from an outside view, hard to get contracts and permissions, and hard to compete with established co's like Boeing that have all of the relationships already.

Check out the SBIR program. It provides an easy way to get into the system and innovate.
Absolutely one of the best ways to get into the system.
Can you comment more about this? Curious about how people get into it. Applying for a grant?
https://www.sbir.gov/solicitation-listing/open

This is a list of the current solicitations available. You can navigate around to the different departments. They have a calendar of announcements and due dates.

If you click on the DoD SBIR[0] it brings up the available topics. Each period has different areas of interest. You create a little proposal and then they announce "Phase 1" awards. They usually have different terms, but I think it's typical to get up to 100k for 4 months. After your demo you can apply for Phase 2 which is more money over a longer term, after that you can move to Phase 3, commercialization, etc.

The problem is that you have to work on it "full time" or at least the principal investigator has to work full time. The other problem is that they give you the topics of interest. There are so many over all of the departments. The biggest are definitely HHS and DoD. There are a lot of ML/AI, IT, sensors, etc.

Unfortunately, the current cycle mostly ends tomorrow for DoD.

[0]https://www.sbir.gov/node/1620805

I think AR has potential to be the most impactful new type of technology, but I don’t have any idea if wide adoption is 1 year away, or 10 years away (or more).

At the moment most of the focus is around overlaying some virtual world in the real world, but I think that technologies like U1 chip being added to new Apple devices could be really interesting to allow objects in the physical world to better mesh with digital world.

Experimenting with live filters for Instagram could be an easy way to dip your toes into AR: https://sparkar.facebook.com/ar-studio/

It's a fun tech, but I fail to see any real world value that can be derived from it. I have worked on AR stuff before and found some niche uses, but I am having real trouble finding general commercial applications apart from fun and games.
There are tons of applications! The thesis behind AR is that your phone (or some other device) can be a passport to accessing a digital world around you. Some ideas, in no particular order:

* Tourism of any kind (think AR museum or city tours)

* Location and event discovery -- as of late there have been quite a few startups trying to capitalize on giving people things to do, but none successful. This is definitely on Snap's radar with their acquisition of Vurb[1] and Placed[2]

* Interactive public art -- Paris has something like this[3] and I really want to see these types of ARGs spread to other cities

* E-commerce -- superimposing clothing or furniture[4] onto the real world

Also, why are fun and games not valid applications? Epic Games, for example, is worth 4.5bn. Interactive media is finally being recognized by the mainstream, and it's just getting started :)

[1] https://www.crunchbase.com/organization/vurb [2] https://www.crunchbase.com/organization/placed [3] https://apps.apple.com/us/app/flashinvaders/id895180919 [4] https://www.amazon.com/adlp/arview

Yep, for tourism. Look at some of the examples this small company is/has worked on: gamifi.co
define "niche".

the construction industry will benefit hugely from AR.

I worked on an art installation that used AR cardboard cards as the "controllers" in a 50+ person platformer game. I also worked on a smaller business card AR project.
I recommend this article from HBR https://hbr.org/2017/11/a-managers-guide-to-augmented-realit...

It actually details how it’s used now to verify ships after construction to ensure that everything is in its right place. They do this by overlaying the schematics on top of the actual ship during inspection. I don’t remember the numbers (read the article a while ago) but the savings and accuracy gains by introducing this were significant.

AR will replace every display in the world. There's nothing special about it, it's just a better form-factor.

You would probably use the same kind of applications, but you'll be able to have 1000 windows open at the same time, located all around your room, easily accessible (just turn your head).

All those gimmicky 'AR applications' are gimmicks. The real use case is replacing our tiny displays.

Imagine your entire workstation as a full AR display that can be any size you want on your desk.

AR is literally real life holograms, there's amazing amounts of applications. Everything you do on your smartphone today can be blown up into AR applications.

Quantum computing is becoming more viable for amateurs to experiment with. A few years ago, Microsoft released an SDK for quantum computing and you can run your solutions on a real quantum computer using Azure.

You'll learn a ton of linear algebra as a side-effect of working with quantum computing, which is the foundation of lots of other fields of computer science, like graphics and AI. I'm confident that the pool of quantum programmers is effectively zero, and I wouldn't be surprised if AI spills over into the field and causes an explosion of demand for quantum programmers. By 2024, it might be a field where you can write your own paycheck.

Microsoft does not allow one to run programs on real quantum computers. Only Rigetti Computing and IBM provide that service.
Unpopular opinion: Domain driven stack.

We have been living in the golden era of software industry, thanks to Moore's law. We were able to afford RISC (= general purpose CPU arch), general purpose operating systems, general purpose languages, general purpose databases etc all because the hardware was going to evolve and get faster anyway.

Now with Moore's law showing signs of death, the future for better computing would be domain driven stack. A quick thought experiment will be that: cloud applications will be written with cloud-friendly languages, using cloud friendly databases, on cloud-ready operating systems and processors that are architected for heavy cloud workloads. Much like how gaming was relying on custom stack for performance (GPUs, play station, X-box, etc)

The advent of TPUs by Google is a symptom of this pattern too. Of course, personal computers with general-purpose-everything will keep existing, but the business industry will start shifting towards domain driven stack slowly and steadily for obvious reasons.

I just wanted to clarify your example. When you say cloud applications, I assume you mean applications written to run a cloud, as opposed to running IN a cloud?

Because if it's the latter, that doesn't sound like a domain driven stack to me.

Not sure what more recent xboxes are like, but the original xbox was a cut down windows 2000 running on Intel and Nvidia. It was really close to commodity hardware and software.
the current Xbox is running a Windows 10 (one kernel design) while the PS4 run a patched up FreeBSD.

Only Nintendo bothers with writing custom kernels, and historically Sony with the PS2 having exotic "Cell" processor units.

The Switch also runs on FreeBSD, not custom.
In terms of hardware, they're running on AMD x86 CPUs. AFAIK there isn't anything special about them, other than having a wider memory bus (they use GDDR5 as opposed to DDR4).
But it had a really thin OS layer, and everybody had virtually exactly the same box, so you could micro-optimize to the exact architecture. IMO it does fit the concept of "domain specific stack", it's just that homogeneity is one of the important properties of the stack instead of unbridled performance.
Re: "cloud applications will be written with cloud-friendly languages"

Carefull, you invest your code base on a "cloud-friendly" language and clouds then could fall out of style. That goes for other components as well.

I feel like the language choice isn't going to be what bites people here, the danger is more architecting to a specific platform, relying on its SDKs and optimizing to a provider's specific resource idiosyncrasies.
That is true, but somebody did mention a "cloud friendly language". Most of the languages in common use were designed pre-cloud. This new thing, whatever it is, could thus have cloud dependencies that current languages don't.
Not sure this adds up for me. So long as network latency and throughput remain asymmetrically limited with respect to machine and CPU cache speed, the implementation detail of data locality will bleed into anything you write. At that point, what makes a language, database, OS anymore "cloud-friendly" than what we have today? I can already get 90% of the way there with Kubernetes, Aurora and any language of choice.
It is still true that most cloud-provider datacenters house racks of commodity hardware? If so, I could definitely imagine a shift to hardware that was designed to support running virtualized environments while keeping power and cooling costs down.

I'm not sure what that would look like. Mainframe-esque, perhaps?

Unpopular answer: Meh.

I see your point, but we are already using specialized algorithms to solve problems on generic hardware (CPU). You can move to different generic hardware (GPU/OpenCL/...) which might be better suited (depending on the problem), or use/rent more generic hardware on demand (cloud computing).

What you're implying is already happening, using/programming "generic" FPGAs to act as specialized accelerators seems to be slowly trending (e.g. Xilinx UltraScale); and if that's working well, "larger" process nodes seem to be getting cheaper these days (e.g. >= 45nm ASICs). But as far as I am aware the tooling and ecosystem for all this is still pretty bad; especially compared to how C/C++ compilers came a long way, JS's ease of accessibility or python's trove of libraries. (Disclaimer: I am not working in that field, so I might be outdated).

So to refine you suggestion: Improving the eco system around hardware synthetization could be a thing?

However, that doesn't seem to be what user richtapestry was thinking of(?).

Crypto! 17% of the worlds GDP is spent on financial services. Software can eat it all.
I think you missed the train friend.
Crypto is the internet in 1996, and you are telling me I missed the train because SUN Microsystems is going to be the first trillion dollar company.
.. eventually going bust and being bought by Oracle?
I think that's the point. Existing cryptocurrencies are very likely to get eclipsed by new competitors that actually solve the problems people face with them. And if you want to have a big impact on humanity, solving those problems is a fruitful area. The financial system that'll pay for our retirements likely hasn't been invented yet.
Crypto is not the internet in 1996. It's a scam. The internet never was. "Crypto is the early internet" is such a lame over-repeated phrase by crypto enthusiasts. Doesn't make it correct in any way.
That's what somebody told me about crypto in 2015, or for that matter about Google in 2009 and computer science in 2002.

There's always another train, right up until the point where everybody assumes that there'll always be another train and world domination is assured. At that point, it's time to hop on another track.

crypto is fake money, and it will never survive government scrutiny.

we need our dollars to be made by governments, not basement dwellers.

All money is fake money, the notion of a currency is a value fiction shared among a group of people.

Most governments are very young, and do not last very long historically. Central banking is a handful of decades old, cryptocurrency is one decade old.

Neither the maximalists nor deniers are right at this point. It remains to be seen when/whether cryptocurrency will replace fiat currency.

cryptocurrency will never replace gold-backed fiat currency.
“The growth of the Internet will slow drastically, as the flaw in ‘Metcalfe’s law' becomes apparent: most people have nothing to say to each other! By 2005, it will become clear that the Internet’s impact on the economy has been no greater than the fax machine’s.” - Paul Krugman
Fiat currency is not backed by gold, in fact, it is not backed by anything.

Also, gold is not inherently valuable. It will become a regular common metal once we shift extraction to the asteroid belt.

The truth is there really isn't much with "inherent" value out there, with the notable exception of potential energy.

Is the Government currency True money?
To borrow a term from Microsoft, "Intelligent Edge." With the shift to the cloud, we are finding there are still many scenarios that should start local for latency, bandwidth, or legal considerations. Organizations still want those scenarios to be easy to manage though.

An example might be in store retail analytics- a set up with a bunch of cameras that can detect what people are touching or otherwise interested in. It makes little sense to ship the video streams all the way back to a cloud provider, but orgs want the capability all the same.

(comment deleted)
How do you create a recurring subscription model without gating the service in the cloud as a SaaS?

/s (or maybe not..)

>An example might be in store retail analytics- a set up with a bunch of cameras that can detect what people are touching or otherwise interested in.

Because the one thing desperately needed in this world is even more surveillance?

This already happens with bluetooth phones in stores. They simply ping you from an array and can track your movement throughout the store.
Most retail stores have been heavily surveilled for a long time by cctv for shoplifting purposes.

One, perhaps naive, hope I have for edge based technologies is to enable scenarios that are actually more private. In my scenario, if the processing of the video stays at the edge and only anonymized or aggregated information is sent up to the cloud/ larger organization, then is much better then if the video is shipped offsite and processed.

It’ll take some interesting legislative efforts to really frame this well, but the Europeans (and the state or Illinois with their facial recognition law?) at least seem to be headed that way.

ML on the edge.
What's "the edge"?
Edge might mean leaf-level IoT devices. Plenty are smart enough to do their own trending or anomaly detection (ie ML), for example, and only report to the cloud the results of that.

So you have edge, cloud, and you might also ponder the "fog", where there's an intermediate aggregation layer. Think leaf devices in some building (edge) with some medium smarts done on prem (fog) before clouding it.

Edge here means the client i.e. smartphones, browsers, or dedicated consumer side devices like the dashcam in the car
(comment deleted)
ML training on phones, IoT devices etc. They then share their learning with centralized ML running on normal servers.
I work in this field so I'm incredibly biased: automated business solutions that cut entry-level data employees out of the equation. You save TONS on the bottom line, and cut out human-driven process that is error prone and difficult to manage. I'm talking about things beyond "API-driven dev", more in the realms of Puppeteer, Microsoft Office automation, screen-scraping (mouse/keyboard), etc. I make API's out of things that other devs balk at - and trust me, it has a lot of market value.

This isn't as "up and coming" as all of the other items people are mentioning, but I'd put it on a "always increasing in popularity" trajectory due to an ever-increasing need. It's not really sexy or interesting, but there will always be a HUGE market for the things that I can do =)

I will warn people that "up and coming" tech is often fad-based and has boom and bust cycles, and personally I'd rather be working for a paycheck then waiting to win the lottery in this regard.

How do you find processes that can be automated? I've often thought that there must be a ton of this stuff in various industries where programmers aren't typically embedded.
You really need to find a domain expert who can at least get you started. I'm working on a problem that I would not have even known existed, if not for running across a couple of engineers who had a business in the field and saw an opening that only a small company would care about (market isn't big enough for the large players).

I know similar situations have to be all around us. The problem, as you say, is finding out about them.

Any midsize company has tremendous amount of things that could be automated. For example, I work at manufacturing company, we produce parts for bigger companies. Bigger companies already have APIs, web portal, etc. We have people who often manually enter data.

Any time, a piece of paper is passed around or when people enter data manually you can either automate or improve the process. People are prone to errors.

Right now, I am rewriting a mission critical application, some trivial changes will save hours a week and improve the integrity of the information. It is not as exciting as writing algorithms but it is nice to see an application written by you used by 200+ people.

Honestly your best bet is to talk to people. I doubt cold calling/knocking on the door of a business will succeed, but people get enthusiastic when it comes to complaining about tedious/monotonous work. Especially at bars. Turn to someone nearby and say "I am looking for ideas to yadda yadda save people time and frustration. Is there any tedious process you deal with at work which you think doesn't need a person to deal with?" If you travel, this is especially common in airports. I've never asked someone this question and yet people rant at me all the time once I say "software developer".
Very common when you do tech work in a non-tech industry. Comes up a lot with coding-inclined mechanical engineers. Zed Shaw's words are really true:

> Programming as a profession is only moderately interesting. It can be a good job, but you could make about the same money and be happier running a fast food joint. You're much better off using code as your secret weapon in another profession. People who can code in the world of technology companies are a dime a dozen and get no respect. People who can code in biology, medicine, government, sociology, physics, history, and mathematics are respected and can do amazing things to advance those disciplines.

https://learnpythonthehardway.org/python3/advice.html

Great quote -- surprised I've never seen it before!

In my limited experience, it's a mixed bag.

Good: you get special treatment/opportunities because of a unique skill set and increased visibility on the end results of what you do.

Bad: management doesn't really know what you do between software releases, you're paid the going rate for your industry while SWEs make far more, and in-house software quality standards might not be established/followed.

Side ish bonus: you're basically prepared to run a one man show or move into niche consulting (much much higher rates). You also get to set the quality standards/procedures going forward, which can be satisfying.

As a tangential bonus, I've accidentally converted my PhD research coworkers to strict git/markdown thanks mostly to typora (windows application). I showed one person what my work flow and version history look like for some internal documentation and now they do the same and convert to word/pdf as a last step. As far as I can tell, no one outside of the math/cs intersection has the patience for latex.

Source: in that boat.

Edit: In regards to market rates, that can be alleviated somewhat in follow up negotiations (6-12 months in or so). It's hard to convince someone what you're worth / what your value proposition is when they're not used to hiring software people. You need to demonstrate your business effect first, since they typically don't have a clear picture of it.

Sounds like a pain in the ass.

It's already a huge pain dealing with know-nothings inside of tech firms. Now imagine having to cater to know-less-than-nothings somewhere else.

I disagree with your edit. Pay is categorically higher in tech than not for engineers.
Man this is me. Knowing a bit of coding and machine learning in engineering has been such a boon over my career in civil/environmental engineering.

But it's like math, if you know it well enough, you'll find ways to use it everywhere. If you don't, you won't. You have to be the type of person that likes to innovate. Its hard to sell to prospective employers, but its great for demonstrating value once you are with an organization. All of my previous employers fight over trying to get me back when I've found myself looking for work. ...Now I work for myself and make my own work and I've priced myself out of their offers, but that's not so bad.

How much machine learning have you been able to pick up and did you learn formally (in school) or just on your own? It's a broad subject, so where would you recommend one begin, assuming I have a decent undergrad math background? Thanks!
Study stats and convex optimization. If you understand logistic regression and MLE, you're mostly there.
Not the parent, but I did an undergrad maths/physics degree some time back and found https://www.coursera.org/learn/machine-learning to be good as an introduction, unfortunately a new job [unrelated] has prevented my finishing the course but I hope to pick it up again later in the Winter.

I would be interested in thoughts from anyone with ML experience who has reviewed said course's materials?

I've always had way too much math under my belt which helps a lot and have taught myself a lot of genuine computer science out of personal interest. I actually did Andrew Ng's Coursera machine learning class all the way through as a first introduction to that field before realizing it wasn't so mysterious and was just the application of a lot of math I already knew, then ran through a bunch of tensor flow tutorials when that first came out and the like. Then just experimented on my own. I have a knack for data though.

Formally from school, I've only had 3 semesters of scientific programming in Fortran and a shitload of math. That and years and years of building models and massaging data in Excel.

Mostly I'm just really used to learning a new API/tool and applying it to new things.

A lot of the ML stuff hasn't been fancy ML, just basic things but applied in really clever and novel ways.

Same money running a fast food joint? You're underpaid bro or you know some crazy overpaid fast food managers. People from science based fields are coming to computer science because those other fields lack job opportunities at competitive pay.
I assuming running = owning, not managing. Subtle maybe in practice but not in revenue.
I believe OP means running a fast food joint as the owner, not a manager employee.

Granted I don't know anything about that business, a quick google search comes up with an article from 2015 saying running a McDonald's provides an average annual profit of $150k so it sounds about in the right ballpark range: https://www.mymoneyblog.com/mcdonalds-franchise-cost-vs-prof...

McDonald’s franchise owners make $500k to $1 million a year, in average. That’s profit, not revenue.

My source is from the McDonald’s franchise disclosure documents. The money blog mentioned in sibling comments claim it’s less.

According to this random quora link: https://www.quora.com/How-much-does-McDonalds-make-in-a-day the average McDonald's unit in the USA has $2,670,320 in annual sales. 25% - 40% profit margin sounds really high for a restaurant, but I don't know enough about the industry to dispute it. Are you sure those numbers were presented as averages and not the high end of what you could make? Or does the typical owner have multiple locations?
My source was this pdf: https://www.bluemaumau.org/sites/default/files/MCD%202013%20...

Very possible I skimmed and may have read it incorrectly, accounting is not my thing. As I have now reached the maximum amount of effort I'm willing to put into a forum comment, I'm not going to dig any further. But if you can tease out better information, I'd be curious to know.

From the document... Across ~12,000 locations they put the majority of restaurants in a range but they do have the numbers pretty well crunched.

Average profit margin 26-28% Average gross sales 2.2 - 2.6M Average operating income before rent/tax: 570k - 716k

"The rent paid to McDonald’s will vary based upon sales and McDonald’s investment in land, site improvements, and building costs."

It looks like that rent paid to McD's home planet is somewhere in the neighborhood of 10% of that investment (yearly? I guess?) but it seems to average about 100k-150k.

Good, this was my takeaway as well. My range was loosely rounded since 716k was something like the 85th percentile.
strongly disagree

1. you _won't_ be happier running a fast food joint - the work is gruelling, and it's so easy to go bust. And you won't bring home six figures

2a. people who _actually_ can code are scarce, even in the tech industry. Source: conducted over 300 interviews

2b. quite logically, your coding skill will be most appreciated and compensated at a Big Tech Co, not at a government department where they will be simply unable to see the difference

>2a. people who _actually_ can code are scarce, even in the tech industry. Source: conducted over 300 interviews

As a young person with an interest in software programming (currently studying chemical engineering but still write C code now and then), what do you look for when trying to find out people who can _actually_ code?

Literally the ability to write working code. You'd be amazed how few interviewees can even put together a working for... loop.
This absolutely blows my mind. I have a hard time even believing it. But having never conducted interviews, I just don't know.
I've heard this story before, but it sounds absolutely insane and I can't begin to imagine (let alone expect!) that such a thing occurs. Is it really true?
a somewhat serious answer: simple things should be easy to you, and hard things possible

examples of simple things: DFS/BFS walks; simple Project Euler problems; or "write a simple game in terminal, maybe with some form of minimax search" (a bit harder), or maybe a parser for simple arithmetic expressions

I don't buy it.

There are tons of software "engineers" who try to escape fintech for FAANGs or even startups, coming from top tier firms like Goldman and McKinsey.

Likewise, bioinformatics generally pays a lot less than traditional SWE roles.

So what I have seen is the exact opposite in practice: people with minimum coding skills in other fields trying to jump into tech firms.

I suppose it's hard to match the salaries at The Big Corps working in other industries. Still, outside of software companies, people think I'm basically a wizard for doing simple things, like automatically generating climographs from JSON files. I just don't get that kind of positive reinforcement from software people. I think it's just more enjoyable to be the hacker rather than a hacker.
That's great, but I prefer to work with people who can teach me new tech knowledge.
(comment deleted)
At the moment many corporates are just automating the incoming invoice process. However, many processes are document (any kind of digital file) based to share information between departments, vendors or customers. Many processes could be automated. To identify a business case for automation worth coding such an application or offer an API we use three main KPI to identify processe worth automating it:

- more than 10 documents per day on year average, e.g a bank will receive new annual reports only in some calendar months but in massive scale

- average number of pages or lines of text per document, the longer the document the more mistakes will be made by humans, as they don't have the time to read everything in detail

- average pay of the FTE who is able to understand and process the document manually should be higher than the average pay of all employees in the company, to make sure the documents encapsulate business value

It's not a fixed set of KPI but helps us to sort out too narrow use cases.

By we I refer to the team behind my startup Konfuzio:

http://www.konfuzio.com/en/

Thanks a lot for sharing such valuable insights!

Small payback : on your English home page under "Operation of the software" > "Information security", there's one too many sentences:

> We set the highest standards both when creating the software and when processing your data. Both when creating the software and when processing your data we set the highest standards.

(First one is better imho).

Is this similar to RPA (Robotic Process Automation)?
Yep - it's definitely that, although I typically avoid the nomenclature because laypeople think "robotics" which is definitely not what I do!
sounds like RPA
I was going to say. Many companies are already investing heavily in RPA (Robotic Process Automation) these days. It’s not widely known in tech circles but kind of common knowledge in traditional companies trying to automate tedious workflow tasks.

When I first saw it, I thought: this looks like a glorified Autohotkey or Automator. But companies like UiPath are doing very well selling this type of automation software.

It sounds mundane but for a company with lots of entry level staff, the paybacks are appreciable.

Coupled with cloud based OCR as a service offerings etc. a lot of data entry tasks can be automated.

> Autohotkey

Yep I've had to employ this software to get things done before. Also OCR doesn't need a cloud vs. a beefy computer - I do a lot of that too.

Yes free OCR software itself is pretty simple and can run offline but if your accuracy requirements are high (for critical documents), different algorithms (typically proprietary) are needed and as well as fallbacks to contracted human labor for verification. At some level it makes sense to outsource to a cloud provider.
> OCR doesn't need a cloud vs. a beefy computer - I do a lot of that too.

Can you elaborate on this?

I've been conflicted about RPA because my view is that if there's an API, you should use that rather than comparatively brittle screen-based "integration". What's extremely compelling about this approach though is that you can cut through all of the friction of API's - even end users understand how to use the UI so they're really empowered to build their own automation in a way that they haven't been previously.
80-90% of what I do doesn't have an API. If there's an API that's functional enough you're darn straight I'm going to be using that vs. Puppeteer, etc!
Most big business use older software, and the small amount of it that does have an API often does not work correctly or is stable. Sad but true. I've ran into a lot of instances where people automate through the GUI because it is the best option. Developers don't realize this because they don't like working with older technology.

Also, at large orgs, software updates can take _forever_. I know of some examples where updating to the newer version of a B2B software taking 7+ years.

And a lot of those automations will need things like OCR or NLP. Even just out of the box Excel integration is a huge time saver. UiPath has activities built in for that. It's more than just GUI interaction. There's a ton of partner technology with out of the box integrations, and you get frameworks to build your automations on, as well as all of the infrastructure and orchestration.

Disclaimer: I work at UiPath

(comment deleted)
Do you have anything going on related to process mining?
YES. I didn't even know that nomenclature before you said it... I employ an ELK and TIG stack for a huge part of what I do. You can very much say that I "process mine" my own process. I also have a ton of regex capability that's often employed to rip through large datasets to extract key data.

Lots of times I have to measure certain things before I begin to fix a problem, and the above stated tooling (typically ran in Docker containers) has really really helped with that.

I have dashboards, warnings, circuit breakers, etc etc. that all help me monitor complex processes, and that help me react to ongoing changes (ex: when a 3rd-party website updates their DOM).

Man.. I wish I knew what you know.

These are the things I need at my current role, and I'm woefully overwhelmed. Tying together processes spread out over multiple systems is a large portion of my role.

I'm having trouble tracking all the various scripts running on different versions of my own packages, and reacting to breaking changes on 3rd party sites. I definitely need a dashboard to tie all these automations together.

I'd love to chat if you're up for it.

Absolutely! Feel free to reach out - you can get to my email address at folkhack.com
Great to hear. I'm finishing my doctorate researching a specialized Process Mining approach and its great to hear there's people doing concrete significant work like that out there
So... automate clerical stuff?

> ever-increasing need

Not sure I'd bet on that -- the point of RPA is to eliminate the job you just described.

This.

I work in a small, physical goods business in the Midwest (~30 employees), and the amount of time spent doing data entry is incredible.

Documents (often PDFs, sometimes excel) flow between buyers, shippers, freight forwarders and internally. These documents are often updated.

I'm looking into EDI for a solution, but there's so much value to be created from cutting out all this data entry. At least 1-1.5 full time employee's worth at our size.

And these industries are so far outside the tech bubble, it's two different worlds.

Facebook and Co. get a bad rap for many things, which I agree with. But I credit these social platforms in speeding up tech adoption rates for the average person, who can now use this knowledge at their job.

I read about EDI in the early 80s.
Right. It's been around for 30+ years and it's meant to solve this exact problem.

However, the cost of training staff and implementation is often out of reach for SMBs, still.

The largest companies in the world - Amazon, for example - still actively use EDI.
(comment deleted)
Ah, I dealt with this data entry manual process many times at work. We were able to build a few solutions to remove 80-90% of the human work for these processes. It took a few years to get good at but now I run a business that specializes in this type of stuff.
And how do you find a solution that will last more than a few years, and also handle changes as they occur?

Software solutions require ongoing work: ongoing maintenance costs of a customised solution could easily be higher per year than 1 cheap FTE.

That's definitely a question to be answered.

More than the # of hours involved, it's the error-prone nature of the work and the lag between when the data should be entered vs when it's actually entered.

These three factors -- hours saved, errors prevented, time lag avoided -- make me think that it's worth automating.

You're right, however, it's not clear cut it's the right way to go.

Yes I did some of this at an old telecoms job. It was incredibly obvious that an automated system could put 5-10 people out of work depending on the sophistication of the implementation. I started automating my own analysis job and left before spreading it or letting it work without super vision. Definitely a lot of low level knowledge worker jobs can be evaporated.
> telecom

That's the industry I'm currently working on... they're generally regressive and have shot themselves in the foot in regards to tech and ease-of-use for a lot of their solutions.

Is there a possibility of ...automating, this kind of automating?

I understand that, like regular application programming, a lot of this kind of automation comes down to requirements-gathering and formal specification of workflow.

But, in theory, could there be some kind of sufficiently-advanced software wizard that does most of your job? Maybe by monitoring entry-level employees' interactions with the software-to-be-automated over some period, and using that as a training input to a workflow guesser?

I'm not imagining that you could be cut out of the loop entirely, mind you, but rather that you'd reduce the human-interaction man-weeks part of your job down to a few hours of pairing with a domain expert to clean up the output of workflow-guesser into a final model.

> Is there a possibility of ...automating, this kind of automating?

Nah - I'm going to be cocky for a second so please take this with a grain of salt... Automation is the only job that's safe from automation ;)

That being said, it's prime for development of tools... I've been working on taking a lot of my workflow out of Charles Proxy and moving it onto a more "man-in-the middle" model that can replay certain web traffic... PARTS of what I do, but automating the automating is probably the most difficult problem that I've had to solve (ie: building my own tools).

They're already automating the exact job you've been describing:

https://techcrunch.com/2019/04/30/uipath-rpa-series-d/?rende...

To expand upon this, UiPath also announced acquisitions of ProcessGold (https://www.uipath.com/newsroom/uipath-acquires-process-gold...) and StepShot (https://www.uipath.com/newsroom/uipath-acquires-stepshot-add...) as a step towards bringing the process documentation and automation planning tasks down to the level of understanding that SMEs and BAs have about automation. These tools will be a really great addition to the ecosystem as I (and clearly they) have found the most difficult work in this space to be understanding the client's existing process and planning to automate it.
This is what Zapier and IFTTT and others are doing.
How do you find customers for that? Any job board tricks, maybe?
Networking and knowing people in industries that have a need =)

I've never really done the job board thing so no idea on that.

Wow, interesting comment. I just started with such a Puppeteer (et al.) project...
Agreed, there is definitely a need for 'smart' automation. There are a lot of tools out there which work however, finding the right tool for the right job is difficult. Also, I think its extremely difficult to get the automation right in 100% of the cases. False positives (even 1%) could be extremely dangerous in cases. Hence, I think there is a huge market for 'smart' automation tools, people who can build them and also for people who can use them efficiently.
One of the most useful programs I made at my last job was an automated letters program.

We had a whole slew of standard letters that we'd send out, most of the time requests would go to the typing pool (yes, this was a thing) that were basically "Do letter ABC for client XYZ", where letter ABC was a standard thing that always went to the same recipient.

I built a screen in our system that let you tell it to request a letter (only the simple letters, obviously), it would pull the relevant data out of the client file and write the request out as an XML file. On the other side of things, a VB6 program would watch a folder for those XML files to appear and go through OLE Automation (as it was called back then) to make Word fill in the letter, print it, print an envelope, and save the result in the client's folder.

This kind of thing is an absolutely incredible time saver.

A whole lot of valuable business software is essentially domain-specific or workflow-specific mail merge.
I've never thought about it that way before, but you're absolutely right.

Another "domain-specific mail merge" that I built was a tool to pull summary data out of the client financial records and populate specific fields in some off-the-shelf income tax software.

That was a terrible tool to write and maintain. The tax software's import feature didn't actually work correctly. I reached out to the vendor and the response was basically "Works for us", so doing it the right way wasn't an option.

Instead, I used VBA to blindly enter data into the program using SendKeys[0]. Maintaining the program was a case of tabbing through and counting the number of times I had to tab. Enter first name, tab, enter last name, tab three times, enter address line 1, etc. Next year when the forms in the tax software changed I'd have to add/remove tabs in the appropriate places, sometimes input stuff in a different order, etc.

As a program, it was horrible, horrible code. As a useful tool for the business, it was incredible. It saved a ton of time and the staff that used to do this by hand, in addition to doing actual work on the more complex files, were incredibly happy.

Overall, it was completely worth it.

[0] https://docs.microsoft.com/en-us/office/vba/language/referen...

> SendKeys

Holy cow why didn't I know about this 6 months ago =|

VBA is something I have had to reluctantly learn. It's a horrid/ancient language. Sorry if someone gets offended at that but to someone who uses a lot of modern development languages/tools it's just very very difficult with near-zero "nice to haves"... you just end up taking this huge lib of code around with you with functions like "inArray" etc.

> As a program, it was horrible, horrible code. As a useful tool for the business, it was incredible.

That's something that's rough for us engineers but can be hedged with great documentation, both inline and and otherwise =)

Do you use AutoHotkey ever?
Yep! Sparingly but for some Windows-based tasks you can't beat it. Also it's a solid solution that has great support/devs behind it so it definitely has a home in my toolbox!
Is anyone aware of a good resource to learn VBA? All the resources I've seen online are either tutorials for very quick hacks, or assume you have 0 programming knowledge. I would love a "VBA for people who are familiar with at least one other programming language".
You don't have to learn VBA to do this, you can just include windows.h in C and call the SendKeys function.
I'm interested in scripting stuff in Excel.
Professional Excel Development Bullen, Bovey, Green
There's also java.awt.Robot – I've used it a lot for similar things, really neat if you already have Java/Scala code for getting the data to fill in.
I’d recommend learning at least one other programming language, then use the tutorials.
I abhor VBA as much as any sane person but this sounds so useful.
Which is actually a terrible step backwards when it comes to interchangeability and iteroperability.

You have data in a database, you flatten it into a non structured document, send it, and then the other end uses data entry to add it to a database. What should be the realm of EDI/CrystalReports/Blockchain take your pick, instead everyone wants to interchange data in the pretty human readable format, as opposed to just rendering it that way for humans while also communicating the raw structured data.

This is sort of like saying it would be a step backward to drop humanoid robots into an office and let them take over tasks. A large part of the problem is that right now, a lot of the interfaces for performing work are optimized for humans, so automation that uses those interfaces is drop-in
If you build things like that, you probably don't have authority or leverage to rebuild everything from scratch, and even if you did, it might be much costlier to disrupt an inefficient web of human-computer-paper interaction that works reliably and that everything else has adapted to.
I've been thinking a lot recently at trying to do this kind of work as well. I'm just not sure where to start. I have a pretty weak network form living abroad for so long and moving around a lot.

Any advice about how to even get started doing this kind of work? What tools to study or learn? How to find work?

Absolutely!

Don't move around all the dang time. I've lived in two places and have an incredibly strong network in both. I find people who are bouncing around 50-times-over usually lack the deep business relationships that it takes to establish trust to solve "big" problems.

As far as getting started - once you start, just never stop. Solve every dang problem you can get your hands on big or small, and continue to publish code every day.

Tools to study and learn - web web web. Also Microsoft Excel automation is an incredible need for lots of businesses. Puppeteer, and learning how to mimic what a browser can do without using a browser would be the two most useful skills I have personally. If you don't have a VERY solid background in webdev you need to get that going first, so much of what I do is reverse-engineering someone's website so I have to be very versed in the way that sites are built (all the way from ancient ones to the most modern of frameworks).

How to find work is network. Once you have the skills, you need to know people that have potential needs. I've already reached out to two people who found me through this thread and regardless of "making money", I genuinely want to help them, if even to just point them in the right direction (like yourself).

People are (*generally) awesome if you put aside your own ego, and I would be absolutely nowhere if I wasn't a sociable person. I love to learn about the world by making connections with people, and genuinely want to help them with what I can do!

I could use help automating a simple task for one of my businesses involving submitting refund requests to USPS. Please get in touch if you're looking for new clients!
Once you create an api or solution, are you able to resell it to other businesses or or do you create once off solutions?
Both - but I'm more tooled to solve one-off situations. I have been able to do the whole "copy-paste" resell a few times which has been financially lucrative!

My GitLab is my toolbox in so many ways and I've got a wealth of clean/documented code that I can re-purpose depending on the task at hand.

Robotic Process Automation (RPA). Examples are UI Path and Blue Prism.

It's finicky technology that is hard to make sustainable, but a lot of companies are really interested in it because of the amount of work that could be cut out like you said.

> It's finicky technology that is hard to make sustainable

This 100 times over. The only way I've been able to make it work is being one heck of a generalist. I've found that training on what I do is near-impossible and I don't even know if I could scale a real company in this beyond myself.

Well, UiPath's valuation is currently 7B... But yeah. They have an "academy" for training RPA developers, forums/ social presence, big sales teams, a process for developing automation, an "automation framework", a marketplace for automation activities, etc.... all these are stuff outside of the core software (robot/"activity executor", orchestrator, and studio for designing said activities).

Maintaining & expanding the automation is a heck of a challenge though. Like, once you got the stuff deployed, make sure that it works without headaches, in the face of other software/ OS upgrades. The robots/automation services are essentially a "fleet of microservices" that you have to maintain for a customer whose strength is likely not IT, and who will change the infrastructure from underneath your "microservices". It's a hard problem, but I feel we're well positioned to tackle that, and if we do, the sky is the limit really. (disclaimer: I work for UiPath, though that should be obvious from the message)

Right on! Yeah - I would love to learn more about what the bigger guys are doing.

Everything you're saying is 100% something I deal with on a day-to-day basis, and ongoing support is a huge struggle to workout with the companies I've implemented software at.

Looking at your product offerings made me chuckle - I definitely have versions of a lot of what you do! My robots are called workers, my orchestrator is called a broker, etc etc ha. Lemme know if you guys are ever looking for a remote automation engineer w/tons of cloud experience ;)

A big challenge with the Blue Prism world is getting the people who will lose their jobs to map the processes Robotic processes will automate. Fiddly, tricky and easy to automate a vast number of errors. Only works once extensively tested and results verified. Ignore analyst and sales hype at your peril.
- Never use an RPA tool that doesn't integrate with third-party SCM. If they tried to roll their own, that's a bad sign

- Never use an RPA tool that doesn't generate plaintext-serialized scripts. You're going to have a bad time if they're binary locked

- Never choose an RPA tool that's been around fewer than 3 years. It's probably just a shim on top of MS automation libs, and can't handle the really gnarly stuff

- Never promise anyone you'll automate 100% of their workload. Never try to automate 100% of their workload. Never hesitate to tell a VP you're not going to automate 100% of their workload. There's value in 50%+, get the easy win and move on. Only come back when you've gotten all the easy wins

- RPA is fundamentally about target selectors (or match rules, or whatever else your tool calls them). Their robustness is the only real feature of an RPA platform, and a smaller toolset is going to result in some fragile, quick-to-break stuff

Ultimately, RPA is about one thing: creating a more tactically malleable layer on top of your existing software. Development and change speed is the biggest advantage.

It shouldn't ever be a core system, but it should be where you prototype functionality.

Everyone listen to this - this is 100% accurate and SUPER applicable. Have reached out via email - thanks for the offer to chat in your other comment =)
Perfect. Also, set project start boundaries to ensure realistic goals and expectations. Under promise and possibly over deliver - all too often data and information discovery reveals unforeseen problems and opportunities
I've got a story about how we were spec'd at handling 50% of incoming workload, hit 60% on the first iteration, customer got so excited that we pushed, and project ended up being canned when we failed to hit the (then) final 95% target.

Taught me a big lesson about realistic messaging and never up-negotiating expectations.

Could Selenium be used for this? It checks most of the boxes you mention.
In my experience, no.

(A) Selenium's UX isn't nearly where it needs to be to upskill an analyst to create their own automations.

(B) Selenium's Windows app compatibility is haphazard.

(C) Selenium doesn't have the kind of corporate support it would take to expand compatibility quickly enough to catch up with its competition.

The RPA space is the Linux desktop problem in a nutshell. Polish and niche compatibility are the final 10%.

Nobody on the open source side has the interest in making a VB6 app work. And nobody on the corporate side really wants to use it for more than what they're currently using it for.

I can't overstate the sheer number of bizarre situations a tool needs to be able to handle to be effective here.

Because being able to automate something 95% of the way to completion is usually a lot less valuable than 100% (note: talking about percentage of happy-path process, not of total incoming workload here).

If anyone in this thread wants to talk RPA, I'm happy to talk your ear off.

At this point, I've been doing it for... about 7 years?

It's been an interesting ride.

Email in profile

Would love to know more. How is RPA different than Selenium?
See reply to your other comment.

Tl;dr - different target user & no corporate sponsor seriously incentivized to improve legacy compatibility

Would love to get the Selenium viewpoint though! Have always been curious, and I'm not as well versed in that side of the house.

Kinda how Dropbox is different than rsync (imperfect analogy, but still pretty good I'd say)
> Lemme know if you guys are ever looking for a remote automation engineer w/tons of cloud experience ;)

Maybe - but how would one go about contacting you? (I have my email in my profile, should you wish to contact me).

Disclaimer: I work for UiPath.

This is a problem with almost any type of automation. I wouldn't say our software (I am only familiar with UiPath) is finicky. GUIs are finicky, but there are ways to deal with them. That's the stuff a good RPA dev can handle.

I was automating before I started at UiPath, and GUI interaction adds a new layer to automation, of course. But it is still maintainable when you implement them using best practices and CI/CD. I didn't have UiPath at my previous job but it would have made a lot of our automations more reliable and more maintainable. We're also making strides to address these types of problems easier.

I strongly believe that UiPath should a tool in any automation developer's toolbox, as well as GUI testing. UiPath is also pretty easy to use, so business users can automate simpler tasks on their own after going through the academy.

The industry is exploding, and good devs are in high demand. Salaries are high, and you can download Community edition and get certified for free.

How is UiPath different than open source tools like Selenium?
It's developer friendly and easy to use. It has an IDE for creating automations. I know a lot of devs might think they don't really care about an IDE, but when you're automating interfaces it makes it _way_ easier. It also is not limited to web browsers.

Check out this video: https://www.youtube.com/watch?v=3B4Y_aUBWTM

It's a very simple automation. It just gets text from a browser and inserts it into notepad. But it's a 3 minute video that does something. Check out Selenium tutorials and compare how much you would learn from a 3 minute video. You don't have to dive into the HTML, you just click on what you want to click/type into/scrape and it knows what you want.

But it still has all the power a developer would want. You can create custom activities using C# and VB. The product itself is not open source, but it is very extensible and flexible. The workflows it generates are text files, which work well in source control. It has source control integration built in, which a custom diff tool. It enables code sharing and encourages code reuse.

The only downside I would point out is that it only runs on Windows for now, which might be a problem for linux only shops.

Any advice for companies getting into this realm, specifically around making sure you get your 'valued' price?

Charging for value can sometimes feel exorbitant... i.e. saving a $100+/hr person ~20 hours a month. Sure I can calculate out it on an excel sheet the customer ROI, but still feels crazy.

My market value is $10-15k/mo USD for just myself (sometimes more if I'm needed in an emergency). If you have a proven track record with a solid reputation you can command that sort of pay without an issue. I also bring a lot of general web development capability to the table, so selling on an easy-to-use interface also does a lot to add to my value.

If you think about it - $35-40k for a 2-3 month dev timeline (which is what I'd say is my average) is worth it if you're cutting the salaries of 5+ people out of the equation, no need to pull out an Excel spreadsheet for the ROI that. All-in-all - you're not crazy, they're the ones who are crazy if they continue to sink money into human robots.

> Charging for value can sometimes feel exorbitant

Just get confident (this is hard advice I know). I'm incredibly confident in the problems I can solve and will put in the hours to get back on a timeline/budget if needed. Businesses aren't buying my "products" per-say, they're buying me.

I owe you a round for this. Hopefully it sticks.
Solid comment. How do you deal with getting the knowledge of the process out of the people you are trying to leave without a job? That must be rough.
They're entry level positions usually so figuring out their processes is typically not that difficult. Often I will not work directly with them vs. having a big sheet of instructions delivered with a ton of bullet points. I also get a ton of process diagrams as well.

Usually it's better for the business I'm supporting to document their own workflows before it's handed over to me for a lot of common sense reasons.

"Don't worry folks, our people are working on introducing basic income, so you'll have opportunity to develop in more important areas of your lives, instead of wasting time on such a meaningless activities." :)
I was working for a scientific journal. In the peer-review department, they had to use this proprietary software, and we had 2 employees who just mindlessly went through this proprietary software's form day-in, day-out. This form spammed scientists for reviews of science articles for our somewhat-predatory journal (we solicited authors to publish in our journal for a fee)...

First, it was a massive waste of those 2 human's intellect, entering data over and over into a generic form from a proprietary business' website, in order to spam scientists. I started a ruby project to automate out all those processes, just using selenium to enter the data onto the web form for us. Before I could get very far, I got a huge promotion elsewhere and left that under-compensating journal.

It all gave me a depressing vibe though, both the ethics of what we were doing and the waste of time/life that those employees had to endure to do it.

> First, it was a massive waste of those 2 human's intellect

This is usually the reality and it's nothing to be ashamed about.

At work, do what you're paid to do. In your free time, pursue interests (literally, things that are interesting to you).

When those jobs are automated, what will the human workers do, will they leave that company? I'd enjoy doing something else if that was my job, but if there was a more useful job there, wouldn't they already be doing it? This is in summary why I think more automation will lead to some people losing their jobs - if there was a more profitable alternative for a company (more productive, paid as well or better?) then wouldn't the company already have automated it?
I worked in the Automated Guided Vehicles industry, and so I want to extend your suggestion: any automation that can replace employees.

Those systems are easy to sell with "This system will pay itself back in 2,5 years, taking into account your current labour cost".

Given all the AI/deep learning possibilities, lots can be automated still.

Do you find it difficult to sell the idea of automating the work of "entry-level data employees"? Middle management generally thrives by building kingdom of such minions.
> Middle management generally thrives by building kingdom of such minions.

I often am not selling to middle management, I ideally sell at the exec and/or product level. Middle managers are a conflict of interest when it comes to what I do.

There exist more powerful forces in the world than "middle management".
Can you give some more insight into what you work on automating specifically?
Anything - green screens, Excel spreadsheets, PDF generation, Docusign, tying 6 web experiences with separate logins into one, simplification of complex auth scenarios, etc.

Most of my day is literally just getting one step further and one step further into a "process" of some sort, and then handling a metric boatload of edge-cases as they come up. Then the last portion of the contract is the actual integration into a business, and handling the "ninety-ninety" rule that always ends up happening when the rubber meets the road.

I take a complex, convoluted mess and turn it into a coherent "something", usually an API or web experience.

That is essentially what I do. I pitch it as process engineering, systems integrator, programmer, etc.
I'm curious how you market yourself or your services? I automate processes, workflows and tasks like this for some clients and I really enjoy it. It's easy to see the value I provide and the clients are happy. However I struggle with how to acquire new/more clients. The ones I have now I've acquired through past business relationships. I'd love to be able to do more of this work.
The key to these types of solutions is always to follow the 80/20 rule. For example, it’s foolish to say and attempt “I will automate all of your data entry people” because there will always be weird idiosyncrasies that are better fit for humans. You will dash yourself against the rocks trying to get a “100% complete” solution.

Instead, create software that allows 1 person to do the job of 5. You create massive business value without getting sucked down the hole of edge cases.

True. In the same spirit, I'd also add record linkage to automated data entry. It's a big problem both if your previous data is noisy or your automated solution doesn't transcribe some fields correctly.

A simple probabilistic programming solution can work really well.

I’m building a tool to speed up data entry. It highlights OCR‘d words with low confidence for human review. It uses templates for capturing structured data and a fast interface for capturing unstructured data.

It grew out of consulting projects and is almost ready for beta testing if anyone is interested in playing with it and giving me some feedback - my email is in my profile :)

How do you guys network with the right business people to speak their language and find out how their processes work? You need that info (a problem description) before you come out with a solution, don’t you?
Just ask. People absolutely love talking about their problems, it'll probably be the most animated thing they have to say about the business they're in.

"Some people profess difficulty at finding applications to write. I have never understood this: talk to people. People have problems — lots of problems, more than you could enumerate in a hundred lifetimes. Talk to a carpenter, ask him what about carpentry sucks. Talk to the receptionist at your dentist’s office — ask her what about her job sucks. Talk to a teacher — ask her what she spends time that she thinks adds the least value to her day."

https://www.kalzumeus.com/2010/03/20/running-a-software-busi...

The "just ask" approach is very tricky. It does not emphasize a critical element : Most people love to rant and complain about things that annoy or bother them. People rarely if ever get to ideas to solve problems or even to describing the underlying problem.
> People rarely if ever get to ideas to solve problems or even to describing the underlying problem.

That's your job. Develop hypotheses about what the underlying problem is and ask them questions to try to falsify them. Develop hypotheses about solutions and build mockups or proofs-of-concept and have them try them out to falsify them.

No one will hand you a business idea on a platter. But problems to solve are the easiest thing to find in the world.

You're right its my job to find solution to problems.

I meant to say 'Just Ask' and 'Problems to solve are easiest to find in the world' are very misleading things for a beginner.

Just asking won't lead you to a solution and identifying solvable problems are incredibly hard.

The question I was answering was: "How do you guys network with the right business people to speak their language and find out how their processes work?"

I don't think "just ask" is misleading at all. The fact that most people will happily complain about their business processes but won't have ideas to solve them doesn't make it "tricky", it makes it an opportunity.

There might be fields of endeavor where identifying solvable problems is incredibly hard, maybe in academia or politics, but business processes? Execution and, if you want to get rich, scaling up are hard. But identifying solvable problems with people's business processes is totally one of the "easiest [things] in the world".

See also: http://www.paulgraham.com/schlep.html

Isn’t this the “robotic process automation” market? (though that one is a bit more than what you are saying)
Isn't this what for a short while was called robotics, but now is mostly referred to as RPA? It sounds like you are competing with things like Blueprism.

These things are really a big deal within large organisations.

Yep! I'm effectively a boutique provider of such services.
This is what I do for work (mostly) as well. The challenge I've had trying to grow the business is "convincing" (for lack of a better word) businesses that they will really save a lot - even after showing them case-studies. Companies can be stubborn.

Hit me up if you ever want to chat! ryan luma.im

I'm living in Ireland and this is what I've wanted to do for years. There are loads of older small to medium businesses with SO many things that can be automated but I don't know how to get into it. I don't even know how to research it. What's the process for getting a client? Cold calling? Partnership with another provider?

What is the job title I should search for more information? Automator?

You should look at business operations analyst, UI Path training is on linkedin learning/lynda.com too. I think information management systems too?
A handful of places around Dublin have active Robotic Process Automation (RPA) teams, which seems to be what's being discussed here.

The Big Four consultancies and Accenture spring to mind. I think the likes of Fidelity and Citi have few roles too.

Note however that you'd be using off the shelf tools to do the automating - you wouldn't be developing the automation tools yourself.

Automation Engineer. It's a broad term, but does carry the idea that you know how to program, and do so in a way that automates normally manual processes. Also doesn't hurt to list yourself on a resume as an Automation/Software Engineer. That gets the idea across that you also understand how software is written, not just how code is cranked out.
How do you find people who need these things?
What’s your business model? Consulting by the hour? Subscription fees? One-off fixed price contracts? Practically all software I’ve ever done has been some amalgamation or other of “make forms better and more automatic”, whether it’s planning software for gigantic Swedish furniture firms (take a wild guess!) or technical analysis toolkits for financial services. It’s always been per-hour or fixed price contracts for me – try as I might but the subscription model simply doesn’t take with the buyers it seems.

What I’d like to do is fixed price for initial dev and then subscription based contract for support and minor updates, but this seems wholly unpalatable for buyers.

Can you share what you're working on? I would love to see a concrete example of something you are referring to as well as how you market it.
I began working in RPA this year and I completely agree. Tremendous potential but also a lot of buzzword-driven hype in the space. I think the main players need to work harder on bringing world-class engineering and best-practices to the community, since the vast majority of developers that are getting into this are not really advancing the field and instead slapping crap together until their 'bot' works. Why are most automation projects relegated to bare-bones budgets and short timeframes? Potential clients i have engaged have a very short-sighted view of automation possibilities in their businesses, and aren't open to the possibilities beyond what they think should be done by a bot instead of their employees.
Currently automating a lot of on-boarding process at work. It's been eye opening. What it's really shown me is how very little expertise we have around all of these solutions, SaaS products, etc... that the company uses. From Tableau to Salesforce to our accounting software, it's a mess. It's put myself and my team into a unique position to gain a lot of mind share, especially when it comes to cleaning a lot of the inefficiencies up. That said, I don't think we're making anyone redundant, or cutting anyone out of the equation in a negative way -- none of the people who are currently tasked with a lot of this work were ever supposed to be doing it, trained on it, etc...
That's exactly what we do, with a nice interface business people can use, email me at rafal@evolution.ai if you want a demo.
I seriously want to begin working in this field as a consultant/architect/developer. So much so that I've registered the LLC, opened a bank account, and even had a few clients I built web apps for. However, I'm not exactly sure how to pivot to business solutions/productivity tools though. Do you usually approach companies where you expect to have a solution for them? Or do you find clients through advertising? Do they find you? Any advice is appreciated!
Does your company, or do any in the field, support policies for retraining/job placement for employees displaced by their technology?
I think UIPath recent success shows that the market already values this.
That's funny, the multi-national company I work for is removing all MS Office automation because MS apparently told us that unattended automation is against the licensing terms. It's OK if you're watching the computer I guess, but not on a server.

Personally this didn't match what I found online (they said it was not supported i.e. if it breaks tough luck) but what would I know. Money must be spent retrofitting other solutions now.

I joked we should pay a minimum wage high school kid to sit in a room and 'watch' the servers but I guess that's too low tech ;)

They have specific solutions that are integrated into things like SharePoint to handle this so they're just putting the hurt on the big guys in order to get more monies.

Total Microsoft play on that for sure - I've never ran into the same issue though.

Yeah I looked into this, and it's definitely overkill for what we wanted (plus sending them more money to fix a problem they made for us would be very grating). All we needed to do is read/write Excel data and similar functions and keeping it local is key. There wasn't much available when this got put in place but there are plenty of reliable libraries now.
I have been working on automating data entry, and the consequences of this is many people (in this case Indians) losing their jobs. I am facing a moral dilemma. I have been writing free software for my entire life, I volunteer, but I live in really poor conditions. I want to get forward somehow, and this seems to be my best chance of getting forward at this point in time and space. I am probably not going to go through with it. It seems to be the case that my principles, my moral values have been holding me back from improving my life. I do not want to lie to people, I do not want to deceive people, I do not want to fsck them over in any way. I do not want to contribute to anything nefarious. But from where I am (the bottom), it is extremely hard to climb up without doing that. :/
Your post brought back fond memories of combining autohotkey with premiere pro to automate boring tasks. Haha. I agree, you have a nice niche.
Automating mouse clicks and key presses seems backward and stupid. Doing this manually will also get disrupted.

You should trace the executable and analyze the function calls. Get a million samples then form a statistical model. The production model replays the function calls given some input. Doesn't need graphical memory anymore and it also has higher resolution inputs (function symbols, addresses) than coordinates on a screen.

Quick control-f tells me no one has said "Environment" or a phrase I'm hoping catches on "Green-tech". There's just no way this doesn't blow up into something big in my mind. Earlier today i was discussing with a biologist friend how there should be a better partnership between the environmental world and the tech world, citing examples of biologists learning to code v's focusing on what they're good at. The environmental fields have not paid well historically, but i have a feeling that if the climate is changing, eventually it'll start paying very well to attract the people it needs.

Stop optimizing driving routes for food delivery, selling CPG products to poor people...start worrying about humanity! It may not pay amazingly, but at least you're doing something that has the potential to be looked back at as important as fighting Nazi Germany or Polio.

> but i have a feeling that if the climate is changing, eventually it'll start paying very well to attract the people it needs

The problem is that it's a tragedy of the commons, so no matter how important it gets, unless governments really start putting money towards it there will continue to be little interest from the private sector.

no disrespect intended at all, but this seems to capture a very non-tech ethos. as an industry we need to have boundless optimism and practice a philosophy of idealism with a healthy sprinkle of realism. we are very well positioned to solve sustainability problems/change culture
I think I partly misread the original post and thought they were looking for a lucrative new field of work; if they just want something worthwhile to put effort into, then I can see how my comment might've come off as cynical. That wasn't my intent.

I'm not saying nothing can be done about our environmental problems, just that it'd be tough to turn it into a career. Unfortunately.

sequester that attitude, dude.

if "governments really start putting money towards it"... that money comes from your paycheck first and foremost.

if you're going to advocate for a market based solution (and you are, if you're talking about putting money somewhere) then consider the hard work of pricing external costs (those costs to the commons, you're so fond of considering), that governments are truly good at.

we don't need governments spending our money. we need governments to help us coordinate prices for the consequences of our desires, which would otherwise not be priced at all (e.g. dumping billions of tons of pollutants, etc... into the shared environment).

> that money comes from your paycheck first and foremost

Fantastic! I would love to help prevent a global catastrophe.

> if you're going to advocate for a market based solution (and you are, if you're talking about putting money somewhere)

I am not. I'm pointing out that this kind of problem is fundamentally impossible for the market to solve unless enormous, non-profit-seeking organizations (i.e., governments) artificially bridge it into the market via coordinated incentives. That's not the only way to solve it, but it's the only way businesses would have anything to do with a solution.

> I would love to help prevent a global catastrophe.

great, you can start by reading books written by folks that don't agree with you.

> I am not.

actually, you are...

> ... into the market via coordinated incentives.

but, like... it's kinda funny.

Build the smallest OS possible that can run Kubernetes at scale in Rust

You’ve just deprecated Linux & C in the data center :-D

Although Kubernetes is for containers, so you'd also need to re-implement Linux in Rust.
Sorry, did the omission of a very detailed list of todos imply I don’t know what this would entail?

You’d need to reimplement a variety of layers and could leave out of a lot too.

Personally a rudimentary BIOS type system in hardware that can read/write to an OS layer would be the design I’d start with

The OS doesn’t need userland tooling then

But no one cares about the OS that powers everything being an archaic mess. Gotta fetishize where the profit is

Climate change will create the first trillionaire.

Find everything that eats fossil fuels and electrify it. Find everything that requires stable land and predictable weather and put it in a shipping container. Find every infrastructure investment that requires decades to pay off and decentralize it.

Instead of water line pipes, pull water out of the air. Instead of copper and fiber optic cables on telephone poles, use solar and satellites. Instead of refrigerated transportation, grow food in your pocket or your stomach. Instead of roads, take to the air.

We won't fix the climate out of kindness. Warren Buffet invested $$ Billions into wind farms because it makes his wallet feel better. Tim Cook just yesterday gave a speech that Apple said "We don't see climate change as risk, but opportunity", that's straight from the world's first trillion dollar company. [1] Elon Musk announced that SpaceX Starship will be pulling its fuel out of atmospheric carbon dioxide on Earth, and on Mars [2]

We'll need to completely reinvent society. Sustainable transportation, vertical agriculture, solar/wind/nuclear energy, air mining, an all-electric economy.

Topics: Direct Air Capture, making products out of atmospheric carbon dioxide, carbon removal. Check out all the companies in the space here: http://airminers.org

[1] https://www.macrumors.com/2019/10/22/tim-cook-talks-sustaina...

[2] https://youtu.be/sOpMrVnjYeY?t=3850

Siberia will be the next California ;-)
Wow! Here's a review of Siberia's climate by 2080: "Even under mild climate change, they estimate a five-fold increase in the potential human capacity."

https://www.newsweek.com/russia-siberia-habitable-climate-ch...

I remember a news crew interviewing Russian citizens on the street about climate change. One guy said, "I welcome global warming, it's f&cking cold here!"
Except permafrost melts into muskeg which is too expensive to farm or build roads on.
Too bad that will be offset by desertification and other negative changes elsewhere in the world.
I shall patent a automatic mozzie zapper
Who's going to pay for it?
perhaps those who wish to survive, after the initial wave of effects begin to be felt
likely individuals, not governments. We bought our way into this problem.
But the number of people who will is very obviously tiny. For any well-off individual, it's cheaper to buy their way out of the downsides. Everyone else is wondering about rent and (in the US) healthcare with a lot more immediacy.

We can see this from the low takeup of carbon offsets.

As one example, imagine that someone invents a process to manufacture gasoline using carbon extracted from the air, at a price that's cost competitive with gasoline extracted from the ground.

Any car running on that as fuel would immediately become carbon-neutral, and it could be produced anywhere on the planet. Almost overnight you could make the entire automotive industry carbon neutral (burning this gasoline would merely return the carbon into the air that was sequestered when the gas was produced) without needing to replace all the cars on the road.

Who would pay for it? Everyone. Individuals would buy manufactured gasoline to power their cars. Governments all over the world would subsidize its production both for environmental reasons (to meet Paris targets), and for energy security to reduce their dependence on oil imports.

That would be a license to print money and could probably produce a trillionaire.

Exactly. Elon Musk announced a few weeks ago that SpaceX Starship will be fueled from atmospheric carbon. They're building a process to pull carbon fuel out of the air on Earth -- and on Mars.

Relevant quote on YouTube from Starship Update, queued up here: https://youtu.be/sOpMrVnjYeY?t=3850

Sure, that would be great. Processes to do this exist - at higher energy costs. The thermodynamic hill is heavily against you on this.

Cheap fusion or safe cheap thorium reactors would also be great. But an awful lot of smart people have bounced off those problems without success.

Yes, once you sort out the industrial process the price of manufactured gasoline would be driven by the local price of electricity at the production site.

Thermodynamics isn't the only thing to consider. There are parts of the world where spot electricity rates routinely go negative, because it's cheaper to pay a consumer to absorb excess electricity than it is to shut off a power plant. A process like this could absorb excess electricity when there's an excess of electricity.

You don't need to talk about fusion or thorium reactors when the cost of solar is low and falling, and we haven't _nearly_ saturated the planets capacity for generating electricity through solar. A combined facility that generated electricity through solar, then either sold electricity or manufactured gasoline (whichever is better in the moment)

There is already a process for it and it's called biodiesel [0]. It has its own range of problems and is not as sexy as the electric cars (that depend on rare earth metals for production of batteries with limited recyclability BTW), but could be much more sustainable long-term.

[0] https://en.wikipedia.org/wiki/Biodiesel

There's a tipping point where it becomes unprofitable to keep existing gas stations open. Then due to lack of available fueling options ICE vehicles rapidly decline. The same thing could happen with financing - as soon as lease payments and average electricity costs are less than monthly fuel costs, the incentive to go ICE will rapidly decline. Especially if range is similar to gas. Not quite there yet.
This could start an ice age
Yes, it could. Actually that's one concern I have looking forward past the current climate crisis: that the current push to solve climate change will spin up a huge industry of carbon capture and sequestration, which will turn the word's economy carbon negative. Atmospheric carbon plummets -- as fast as it's currently rising -- and we put the Earth into an ice age.

Basically humanity has reached the point where we need to learn to regulate our global carbon emissions to keep the atmosphere at a steady state. We're just now starting to figure out how to down-regulate our carbon emissions; after that we'll need to figure out how to up-regulate it in a controlled way.

You either pay to fix the problem, or you pay treat the symptoms with insurance premiums.

You pay no matter what.

FED money printing and the subsequent bubbles will create the first trillionaire.
Zimbabwe beat them to it
Good point. I am a 100 trillionaire in Zimbabwe dollars.
If anyone wants to start a business selling suitcases of money from hyperinflation countries, I really really want to buy one large suitcase full (preferably money with English writing from a known county, Zimbabwe seems a likely candidate!). Or perhaps prop money (although it needs to seem like a realistic Western currency). Or maybe zero dollar banknotes?
You mean geo-engineering specifically, or just a transition to a renewable economy?

Olivine weathering and ocean cloud seeding both seem like reasonably likely solutions that each cost on the order of 10's of billions. Major doubts that addressing carbon will be a big money maker.

But obviously the carbon economy is huge, so the renewable economy will at least that big.

In terms of "What's a Promising Area to Work On", I'm excited about all of it. I've expanded my description to include the renewable economy as well as the carbon economy.
I don't really see that happening. Most of the things you've suggested would require large amounts on energy to produce. I don't see energy getting cheaper over the next 50-100 years. It seems to me that climate change and reduced dependency on fossil fuels we lead to reduced globalization and population reduction.
The cost of energy will go near zero. The marginal cost of an electron captured with solar is free.

Solar is starting to beat fossil fuel installations already, below $0.05 USD per kWh.

This was not predicted even by the biggest cheerleaders even 10 years ago: "In 2017, the solar industry achieved SunShot’s original 2020 cost target of $0.06 per kilowatt-hour for utility-scale photovoltaic (PV) solar power three years ahead of schedule, dropping from about $0.28 to $0.06 per kilowatt-hour (kWh)." [1]

Sunshot goal for 2030 is $0.03 per kWh.

[1] https://www.energy.gov/eere/solar/sunshot-2030

Someone is going to cover the Sahara with solar panels and make some serious cash. Bonus points for automating the production of solar panels out of sand.
Call it "Sand to Solar". Some of the best ideas start out as jokes, I googled for this because I was curious:

"New technology allows heated sand to generate electricity, presenting a viable new option for investors to focus on."

https://oilandenergyinvestor.com/acq/new-energy-from-sand-wh...

I wasted more time than I care to think reading that article. There is no mention of how the technology works. A vague reference to an academic project in the UAE. Most of the article is just filler fluff. This is a marketing article to investors for this authors "wealth building energy advantage".
Okay, how about the copper that transports that electron to somewhere it can do useful work? How about the 1800C furnace that melts sand into the silicon for the solar panel? How about the diesel fuel that the massive mining trucks used to haul the copper ore out of the mine are fueled with (these things are heavy and get something like 8 gallons to the mile (not a typo) good luck running them on batteries)?

Those things aren't going to zero anytime soon.

This is just pointless smug contrarianism, I hope you're just in a bad mood and don't really believe or think this way regularly. Mining trucks are excellent candidates for battery electrification, Volvo is already building and selling them.

What other things do you think are never happening that are infact already happening?

I really doubt it makes sense to use electric vehicles for mining, but I could be wrong. I guess we'll be able to see in the coming years what fraction of mining vehicles are electric.

But what is wrong about what I stated? It's obviously impossible for solar to approach $0 because of the large material costs inherent in manufacturing solar panels and transporting electricity. This is not contrarianism, it's realism. We probably use more energy per capita than is sustainable, and eventually we will have to change that.

Also solar panels wear out. I could see a case being made for the cost going to zero if you amortize it, but you can only amortize it over about 25 years, after which you have to replace the panels. This sets a floor on the price.

> I really doubt it makes sense to use electric vehicles for mining, but I could be wrong. I guess we'll be able to see in the coming years what fraction of mining vehicles are electric.

They're slow-moving, inherently massive and heavy, and travel relatively short distances per round-trip. So they don't care much about battery weight, size or limited range, and appreciate the massive torque from low RPMs.

It's kind of an ideal case for a battery-swapping BEV system. Just wait for the battery costs and energy densities to both improve and this will be a no-brainer. They can charge the drained battery on the grid from cheaper/cleaner sources while the other is busy hauling a load.

Of course if the terrain is such that the mine is on the top of a mountain ascended empty but descended full, you don't even need battery swapping, it'll recharge on the descent.

Nobody wants to spend money owning and operating those diesel engines if they don't have to. When the BEV option is available and makes business sense they'll switch immediately. They'll require little maintenance, and you don't need specially trained technicians to swap batteries and order new ones when they need replacing.

While I wanted to agreee it's just contrarianism, he has a point. Mine electrification is going to come very, very slowly. There are 2 principal reasons,

1) mines run on a 24/7 schedule. There is not enough down time to charge batteries in a shift and with how completely filthy machines get we need better solutions for battery swapping in mining before that becomes an option, and

2) because (like the mine I am at right now) a non-negligible number of mines are far from grids or clean power sources, a lot of battery power would come from fossil fuels burned near the mine.

There is interest, but most of the actual use seems fairly superficial (public image boosts).

Transporting energy long distances seems like such a waste. There's the cost of transporting fossil fuels or power lines for electricity.

What economics would it take for your mine to have its own solar or nuclear powered microgrid. What about when fossil fuel supplies are unreliable or interrupted.

Appreciate the thoughtful questions, and inquisitive responses! Great to see people figure out how we might push these closer to zero. We'll need to lower the costs of manufacturing solar panels and transporting electricity + raise the lifetime of panels.
> Mining trucks are excellent candidates for battery electrification, Volvo is already building and selling them.

For near zero? Where?

Mining trucks are going electric. They actually make energy when loaded and going down a hill.

https://hackaday.com/2019/08/22/electric-dump-truck-produces...

Furnaces have been run off of solar (parabolic mirrors) for quite some time.

While it sounds like an excellent candidate in all mines, one must keep in mind that the majority of mines tend to move ore/waste rock upwards. In those other, rarer cases (for instance: mountain top mining), electrification of the haulage fleet makes good sense.
Good to know, thanks! Seems obvious this is the way to go. Excited to see growth in this space in the future.
> How about the 1800C furnace that melts sand into the silicon for the solar panel

I wonder if a process similar to that used for molten salt batteries, using reflected, concentrated sunlight, could be used to melt sand into what is needed for more solar panels, electronics etc, and use the stored energy while it cools off, essentially combining the processes? And of course for other [s]melting processes powered by fossil fuels today?

Hmm, after doing a bit of research: the melting point for saltpeter (used in molten salt storage [0]) is only 550°F, while silicon's is 2,577°F... would a reflecting solar array be able to reach those temperatures? If I'm reading this [1] correctly, you could only practically get to 3,698.33°F (really close to your example of a 1,800°C furnace) if you collected all the sunlight falling onto earth. It's got to be much worse than that though in practice, because to focus all of it onto a single point would require beaming reflections from the perimeter a long distance through air, and also around the curvature of the earth...

So, we need to build this on the moon or in orbit? Oooh, Futurama actually showed us what could go wrong here [2].

[0]: https://en.wikipedia.org/wiki/Solar_thermal_energy#Molten_sa...

[1]: https://en.wikipedia.org/wiki/Concentrated_solar_power#Ideal...

[2]: https://www.youtube.com/watch?v=0qksm5cRtcU

Why not just use an electrical furnace and power that with solar power?
The furnaces for making crude silicon are already powered by electricity. So are the Siemens process reactors for refining silicon into solar and electronic grades. So are the Czochralski crystal pullers used to turn refined silicon into mono-crystalline boules for manufacturing wafers.

"Silicon processing: from quartz to crystalline silicon solar cells"

https://www.pyrometallurgy.co.za/Pyro2011/Papers/083-Xakalas...

> Those things aren't going to zero anytime soon.

Not with that attitude

even if there were tens or even hundreds of UNIVAC installations, who’s going to run the telegraph lines for all the teletype terminals this would require??
> Solar is starting to beat fossil fuel installations already, below $0.05 USD per kWh.

I read that over and over, but it's like comparing apples and oranges.

(comment deleted)
Depends upon your country.

Countries with significant lake hydroelectricity (not run-of-river) can "store" the solar power (no batteries needed) by reducing flow during the daytime (and increasing flow at nighttime if required).

This is because hydroelectric dams are essentially stores of electricity. No need for pumping or other expensive storage schemes, is your already have one!

Does your current price of PV include storage?

Also the PV cells made mainly from China are currently manufactured using coal power plant. But what would be the price of photovoltaics if only renewable energies were used? My guess is it will be much higher.

I have never understood what people are trying to argue when they make this point. It would be like an oncologist arguing against surgery that would inhibit a tumour on the basis that it wouldn't eradicate it.
Hey, I work at a nuclear fission company. We think we can get our electricity prices to around $0.0025 / kWh by around 2045. This represents a ~20x reduction from off-peak cost in Ontario where I live, and ~150x reduction from average electricity prices in Germany.

Also, w.r.t. population, the way I see it our population is growing very quickly whether we like it or not (look at Africa). I think there's a good opportunity for us to support a population much larger than we have right now with improved sustainable energy tech.

(Solar panels are also getting cheaper, and will probably run around $0.02/kWh in the near future if iteration continues, and are probably capable of supporting us up to 20 billion people as well)

Hmm, Hinkley C is signed up at a strike price of £92.50/MWh, in 2012 Pounds. A 400x reduction when so much of the cost of nuclear is safety, waste management and decommissioning? That's surprising to say the least.

Can you give any clue on how?

The trend of solar and wind continuing the inexorable march downward seems much more certain.

· Liquid fueled MSR, safety is best of any other source. So, we think regulators will lower barriers over time as we prove this.

· Waste management is easier since our waste will be ~5x less mass per kg fuel (~500x less mass per kWh energy), and also returns to natural radioactivity much faster (one or two centuries). We actually plan to sell most of the fission products as useful materials.

· As of now, our reactors are planned to be mostly stainless steel. Most of the lifetime mass-throughput of current GI-GIII fission plants is concrete and steel, of which we have basically none of the concrete and much less steel because no pressure vessel. So, decommissioning is much easier and probably will be internally profitable.

You're right about the last part, because solar will almost definitely get at least 2x cheaper than it is today, and within not more than a few decades.

My perception is biased (ofc), but in my view it's actually very likely that we will reach or surpass our goal, though, so I personally wouldn't bet on renewables.

Thanks. There's been quite the resurgence of interest in salt since it just about disappeared in the 70s or 80s.

I hope you manage it - though I tend to think we'll have a mix of sources rather than making renewables irrelevant. That may be reserve from having heard the "too cheap to meter" slogan a time or two too often. :) Even properly competitive nuclear will make decarbonising far easier, as I'm not at all convinced by grid scale battery, so I hope someone manages...

Yeah, I agree with you completely.

I think we will probably get fission to like 4/5 of the world energy supply and 95% of electricity within our lifetimes.

Solar panels are a great thing to have on your house if you can afford them and want the security in the case of some kind of grid problem. They're also good if you need energy in the middle of nowhere. We want to put fission reactors in most remote communities, even small ones, but if small enough groups of people are going out into uninhabited places it doesn't make sense, so solar panels are better.

I don't really see any utility in wind energy at all, other than in areas where solar doesn't make sense.

Batteries are a really environmentally bad idea for non-transport/device energy supply, because of density. So I think it's best we avoid intermittent sources for most of the grid power.

I'm excited about nuclear, but it seems like we will eventually run out of uranium fuel unless we switch to breeder reactors. Am I off on this? This is assuming that nuclear usage grows to become a larger portion of our total energy use. I see nuclear as potentially solving a lot of our problems, but it seems to have some difficulties as well.

I'm more skeptical that we can support such a large population. Right now we make heavy use of fertilizer, which is produced from natural gas and by mining phosphates. Probably we could avoid using natural gas, but this would require using more energy, so I'm not sure how sustainable this is long-term.

You're absolutely right, U235 reserves are only enough to last a couple hundred years at current consumption and exploration. We could probably sustain a long time with burner LWRs but it will require massive investment.

Breeders however (whether you go with Th-U or U-Pu fuel) both should be able to take us to at least another 10,000 present-equivalent years on a wholly nuclear powered economy.

You're also right to suspect fertilizer, I feel the same way. I've seen some of the phosphate mines around the world and it's kind of horrifying, so I'm hoping asteroid mining can fix our metal extraction problem.

I heard C-Type asteroids have plenty of phosphorus, but I'm definitely not an expert on asteroid mining. If they do, though, it will likely be enough to sustain us for quadrillions of human-years (humans * years) because the asteroid belt is just so damn massive compared to the Earth's crust.

Wow that's really exciting, I need to pay more attention to nuclear.
I'm pessimistic about governments coming together and coordinating an expensive effort to avert climate change. I also think the world is on a course to become increasingly divided and rivalrous, such that collaboration would be less likely, and nations less willing to compromise on their economic competitiveness.

I think if there is an effort made (which there will probably have to be), it'll be something cheap and dirty, like stratospheric aerosol injection.

Same here regarding govts coming together. I think individuals have a huge role to play. The rise of the climate hackers.
The reason there seems to be so much rivalry today is because only the shitty, attention grabbing and warmongering leaders and nations get all the attention.

In the meanwhile, you have most nations around the world living and making progress peacefully. Nations that can be easily convinced by some form of aid or assistance to get on the bandwagon.

Don't be fooled by the news media. Most peoples of the world just want to live and let live. However, if the leadership of the US continues to be dominated by anti-science racist right-wingers, I am sorry to say, this won't happen. That is probably the biggest threat to any progress, not inter-government cooperation.

> anti-science racist right-wingers... is probably the biggest threat to any progress

don't really know what race has to do with it, but big international banks are a bigger threat since they won't let any nation unplug from the global economy. follow the money.

Racism is a strong indicator of lack of integrity and a belief in the scientific process. If you believe that people are less intelligent because of the color of their skin (despite numerous scientific evidence failing to show any correlation between race and intelligence) it’s a good indicator that you don’t believe in the scientific method.
It won't be governments that avert climate change, it'll be individuals acting in their own self-interest.

This might be violent (everybody kills their neighbors, reducing the population and hence greenhouse emissions), or it might be economic (as fossil fuel prices rise, renewables become increasingly price-competitive, such that some people end up making a lot of money electrifying the world). Personally I'm rooting for the economic solution.

> Personally I'm rooting for the economic solution.

An impossibility because so called 'economic growth' (which is no more than global entropy increase) is coming to an end as degraded ecosystems worldwide fall into collapse. Climate change is only the most salient of literally thousands of distributed causes of this, but the overview is: the physical loans taken out on our global systems over the past two centuries are coming to maturity. It's payback time.

That AirMiners site takes 30 seconds to load.
Hmm, wonder why that is. Loads fast for me!
This is interesting to me because it’s such a cultural blindspot: the left-wing point of view is that energy efficiency is the only way forward, and the right-wing point of view is that more traditional fuels is the only way forward.

But solar panels seem to have a version or Moore’s law - and even a single doubling of efficiency from here would completely revolutionize our energy economy (and dramatically shake up world politics!). Two doublings and we have a green future. Three or more doublings and we suddenly have the biggest energy surplus in human history.

It might be a little slower than the computing revolution - but it also might not be. If this is 1980, and solar is at the Apple II phase, the next 30 years are going to be wild.

Solar panel efficiency does not have an analogue to Moore's law, and their efficiency physically cannot double two or more times.

Solar cell record efficiency is 46%, that only gives room for 1.12 doublings theoretically, and thermodynamics is nasty in the real world so we probably cannot even get past 1 doubling from here.

You are correct about thermodynamic efficiency. I interpreted it to be about economic efficiency. In which case one or more doublings of kWh output from dollar input is still a stretch but plausible.

Modern rooftop solar modules haven't even doubled efficiency from 40 years ago, but they have improved cost per kilowatt hour by more than 100x.

Yeah, I tend to agree with you here. The difference I see is that solar has already been heavily invested in and developed recently, so it's probably near the limit of today's material science.

My personal philosophy about technological progress is based heavily on the idea that material science is the principle barrier to what we can accomplish. That's why humanity's progress is recounted in "ages" named by material.

So, I think solar panels can get maybe 2-10x more economical, whereas other options like new hydrocarbon sources, fission, and fusion can probably all get at least 10x more economical than the maximum possible with solar, purely due to the physical limitations of the technology.

I can believe significant price improvements to come for solar, fission, and fusion since their fuel costs are zero or tiny. What new hydrocarbon sources could be 10x cheaper than coal or natural gas?
Well, I'm not really supposed to talk about this but we're working on a hydrocarbon byproduct of our fission reactors. Can't extrapolate but point is, there are unconventional ways to make hydrocarbons that become economically viable when they are a byproduct of an already profitable process.
One thing I've never been able to figure out is that given Crispr, why it wouldn't be possible to engineer an organism that can suck much more CO2 out of the air than any existing tree / algae could. Even more ideal if we can make a good building material out of the engineered plant to sequester Carbon.

It's easy to see how the manual selection process would work here - plant a bunch of trees, see how much O2 each produces and pick Top N and repeat. Of course, this needs thousands of years to run manually because the improvement with each generation would be minimal.

Is the issue that we don't know what gene(s) control photo-synthesis?

Not my area of expertise, but I would assume that there is an efficiency limit, much like there is for solar panels.

Remember that organisms can't just pull CO2 out of the air for free. Plants do it via photosynthesis, which requires the sun's energy. There is a maximum amount of energy falling on each organism, and I would guess that there is some theoretical limit to the efficiency of photosynthesis that is well below that limit.

Yup I agree with you almost entirely. The only thing I was thinking of was if evolution would optimize for the amount of energy a plant receives vs other things like survival, reproduction etc and how much efficiency we can gain by optimizing purely for photosynthesis.

Also I don’t know if plants are adapted to the high CO2 atmosphere yet so even if evolution optimized the ability of a plant to photosynthesize, given this happened in relatively low CO2 environments like forests and without man made climate change, it’s possible that human intervention can improve it.

>One thing I've never been able to figure out is that given Crispr, why it wouldn't be possible to engineer an organism that can suck much more CO2 out of the air than any existing tree / algae could.

Jurassic Park Effect.

Just because we can doesn't mean we should. CRISPR is pretty unknown and releasing man-made organism we don't know every little thing about could topple the ecosystem incredibly quickly. Not to mention how incredibly divisive it would be to do so.

trees have been optimizing that problem for a long time already, and then if you don't bury the tree and prevent decomposition, you don't really solve anything but create a short term buffer.
How long for a "full" tree to decompose?
using the lumber in high rises (through gluelam construction) is cheaper than steel and is carbon negative.. virtuous cycle indeed.
The answer is pretty simple: carbon fuels release energy when burned. This means an organism reversing the operation needs to absorb energy (in the case of photosynthesis, solar energy). Sunlight is not that energy-dense, and concentrating that energy too much would burn the plant anyway.
This is being worked on. The problem is we are terrible at engineering biosystems, CRISPR currently works mainly to knockout genes rather than the kind of enhanced evolution you describe, and upscaling a solution to the point it can have a planetary impact is non-trivial.
This is the correct answer. We are abysmal at engineering biosystems. As Stewart Brand says "We are as gods and we must get good at it".
Re: upscaling, I was reading another article on HN about a specific species of plant when grown in sufficiently large quantities (6x size of Texas I think) would basically bring us down to carbon neutral. We can probably scrounge up that land if the world works together but who knows.

Re: engineering biosystems, I imagine engineering native plants (if possible) has much fewer ecosystem wide chain effects) than using gene drives to eliminate mosquitoes or whatever.

Re: crispr, I see photos of glow in the dark monkeys and super muscular dogs, so as a layman it seems like cutting a gene and vaguely throwing the desired gene near the cut location seems to work. We don’t have as many ethical implications about rapidly iterating on plants as we do on animals, so it seems like if an experiment doesn’t work, just repeat 1000x until it works would do it. People talk about off-target effects etc. but you won’t really know unless you try I guess.

There's actually a tree that does. Paulownia. Reason is that it is also the only nitrogen fixing tree, so it can grow at a much more rapid rate.
As I understand it, algae growth is generally limited by iron, so just removing the limiting factor can increase growth and pull more CO2 out of the air. There has been at least one rogue geoengineering project along those lines - https://csi.asu.edu/ideas/the-first-of-the-geohackers/

Maybe we don't actually need genetic engineering - maybe we can just build big algae ponds and feed them iron supplements. Basically build algae heaven, open source it and random groups across the globe can start doing it.

Discontent amongst world populations is growing, with wildfire riots increasingly breaking out. The root cause of this discontent is the gross inequality encouraged by (amongst other forces) contrepreneur culture. The near-term world is not going to be hospitable to the uber-rich, fortunately in my view. The world's masses are coming after them.
You wish.

Just like the early socialists of 19th century wished.

Many of his suggestions are outright harmful, like pulling water out of air. Same thinking got us in this mess in the first place, hey just pull out oil of the ground and burn it.

Oh yes and no I guess. A certain Schadenfreude would ensue for sure. But as I'm pretty sure it's too late to save our civilisation now, it'll be a lot of chaos & suffering for not much.
Is climate change really tied to wealth inequality or are you just looking for a way to rationalize your blood lust?
Well it's tied in one direction: poorer populations will suffer for the material benefit of richer ones. My nation is quite explicit about this - our government is on record as being happy for Pacific nations to lose their land as long as Australia continues 'economic growth'. And we'll even allow them to labour for us in exchange! The other direction is conceivable: more equal populations may be less liable to vanity status purchases, and less resistant to equitably-distributed change. Perhaps hard to prove either way.

Anyway, that's not much to do with my comment which was merely noting that disaster capitalists trying to make trillions from climate change may meet mass resistance.

As for my alleged blood lust - well I doubt the discontent will result in many billionaires being torn limb from limb. Though as many seem to think paying their fair share of tax would be pretty much equivalent, why not?

Naomi Klein’s most recent book has parts that argue that addressing wealth inequality is a must if we want to address climate change.

In her words, wealth equality is tied to a “climate justice” movement. Some points:

- If you tax the struggling lower middle and working classes to raise funds for climate action they will get justifiably angry that they’re being stressed when they didn’t really create this mess and have the lowest ability to pay. Eg. Frances ‘Yellow Jackets’ - The antidemocratic forces that are produced by wealth inequality have allowed fossil fuel industries to maintain the status quo despite popular support for climate action

I would argue the ones lusting for blood are those who, like decision-makers at Amazon, create abusive conditions for the people who work for the companies they manage.

I'm specifically referring to the disgraceful way workers are treated at Amazon shipping facilities.

Most blood seems to be spilled by those with money wanting more.

One could argue that the arms business spills more blood than everything else put together ...

> Find everything that eats fossil fuels and electrify it

You start off with a free lunch fallacy.

Open an led bulb sometime. You’ll find a large heat sink, at least one fiberglass circuit board, 30-100 electronics components with their own printing / cases / production footprints, ROHS-exempt parts have heavy metals like lead and mercury, you’ll find metal stampings and plastic coated wire, there will be plastic lenses and injection molded mounts, there will be LEDs themselves of course which have tens of thousands of hours time into their design that come from specialized facilities... all the production and shipping and testing and development that goes into each bulb.

... remember old light bulbs? Glass, a wire, a metal thread.

Remember the argument that LED bulbs will last 10 years? They don’t. Tell me... which one do you think will be better in a landfill?

I still think it’s the right move to go with LED bulbs overall - but I’ve been around the world, I’d like to think I have a sense of scale.

When I see people write “just make it electric” i think of things like but not limited to my bulb example and I’m confident that person doesn’t understand much on making things at scale.

Never said it would be easy. But fossil fuel powered light bulbs aren't the long term solution, and renewable-powered bulbs definitely are.
> Climate change will create the first trillionaire.

> Elon Musk

There you go...

> Climate change will create the first trillionaire.

Never attribute to a changing world that which can be attributed to inflation.

>Instead of water line pipes, pull water out of the air.

I thought this story was put to bed a long time ago? [0]

[0] https://www.youtube.com/watch?v=LVsqIjAeeXw

Check it out, this is a different implementation that works!

"A new device that sits inside a shipping container can use clean energy to almost instantly bring clean drinking water anywhere–the rooftop of an apartment building in Nairobi, a disaster zone after a hurricane in Manila, a rural village in Zimbabwe–by pulling water from the air."

https://www.fastcompany.com/90253718/a-device-that-can-pull-...

This is an ideological statement, which I do not believe is based in reality.

There have always been 'meta causes' of various kinds, and they didn't create billionaires.

If there were something very specific about the sector that created such conditions one might be inclined to agree, but I'm doubtful.

'Climate Change' is like 'AI' in that it will affect everyone, everywhere, in every industry, but there might very well be very few 'AI companies' that are huge, in much the same way there may be very few 'climate change' specific companies.

Consider that the 'climate change' movement is well afoot and has been for quite a long time ... and where are the billionaires?

Oil rights can be acquired at some scale, solar and wind, not really.

If someone advanced some nuclear tech, lobbied to get approval to build a lot of cookie-cutter plants built around America, had major subsidies, huge protective moats (i.e. $10B entry point to get in the game) tons of IP and know-how - like the 'Space X' of Nuclear - then maybe we could see a billionaire there.

But I don't see any companies or entities on the horizon, or on the theoretical horizon that would validate this claim. I think people might think it's true because the 'want to believe it's true' but that doesn't make it true or even more likely.

That said ... it's probably a 'promising area' to work in.

Figure out a way to apply modern software automation to a root problem of society. For example, hunger.