Ask HN: What non-obvious tech/market may take off in the next few years?
Hi HN, I'm trying to find some exciting but not-yet-mainstream tech or market to look into.
For the last couple of years I've been completely focused on one field and have not been staying on top of what the latest hot tech trend is, so kind of lost what I should be looking at. Note that I'm not just chasing some tech hype, but just want to know what I've been missing out on.
This doesn't have to be brand new tech per se, but could be a re-application of a previously failed technology which makes sense now because the world has changed.
Please share anything you think is really cool that may take off soon. Also would be nice if you shared the reason for why you think it will be the next couple of years when they take off to mainstream. Thank you!
175 comments
[ 4.8 ms ] story [ 213 ms ] threadThere's a cool project trending on Github right now showing how magic-like some of the technology in this space is: https://github.com/cyrildiagne/ar-cutpaste
https://www.ptc.com/
Apart from gamers, people just don't want it.
Also I would say "apart from gamers, people just don't want it" is the same reaction as people who didn't understand what a "horseless carriage" would be used for when it was invented.
This might be somewhat niche area though.
Reasons this sector might take off: recent greater consumer spending on video games (especially with the pandemic) & their normalization as a field of entertainment, recent greater consumer spending on online services where you pay to interact with actors (although they're all of the, er, amorous type), and the growing popularity of mixed reality driven by the release of Half Life: Alyx.
* No knowledge of acting or actors; never even been to LA
* Never worked with mocap technology
* Never worked with video game development
* Never worked with voice or streaming services
Maybe I can pick up some of this as a hobby. Anyway it's not really a novel idea (like I said, you can read this exact concept in The Diamond Age, a very popular novel among the tech crowd) so I'm sure someone is trying it. Who knows, we'll see in a few years!
1. As you said, the costs involved. You'd probably need to charge people something like $10 per interaction with a voice actor, plus $2 per minute. This makes for a very limited market.
2. Actors need to be waiting for a request on a wide range of characters since there would not be enough demand for them to consistently play the same character all the time. So, when a player triggers an interaction, there's going to be a delay as the actor is brought up to speed on their role, the person they're interacting with, the world they're in, etc.
3. When players return to a NPC, they'll likely be assigned a different actor since the previous one is either in another call or not working. This kind of ruins the premium interactions since the player remembers the voice and small details of the last conversation, but now the NPC has suddenly changed.
4. You're in the business of matching up adult voice actors with children playing games. You'll need a record of every single conversation because it's only a matter of time before you get reports of inappropriate conversations unfolding.
5. This isn't a long term business. Technology gets better each day at speaking and understanding human voice. It will replace the voice actors and it also solves all of the above issues.
Purchase a license to use an actor’s voice/likeness. Have them complete a training set. Now you have their digital doppleganger who can act almost any role required of them.
I could imagine having a very small number of paid actors for key roles, maybe the two opposing kings of a region or something, but eve online doesn't need to pay the heads of corporations, and plenty of interesting stories emerge naturally.
People using apps like cashapp to ditch traditional banks leading to loans and other financial activity to be done over these newer apps.
https://www.glofish.com/
I don't doubt we will see CRISPR and friends applied to ornamental plants the way they are already applied to food crops.
The problem these tools solve are more related to infrastructure rather than one's ability to code. Learning to code might be the easiest part but deploying it, maintaining it, scaling it, securing it and integrating it with thousands of other services remains a huge task even for experienced folks. It's wasted time and effort for something that is cookie cutter in functionality and limited in scope (vast majority of web).
Some of them provide collaboration tools, development environment and experts on call which is neat.
Add ease of outsourcing, too. The employer doesn't have to worry about maintenance once the product is finished. Many good platforms will allow you an easy migration path and better security controls. That comes at a vendor lock in. That's the price but given the life expectancy of smaller companies and startups, it may as well be worth it.
Not sure what the equivalent in no/low-code would be.
It's one of those things where a difference in degree in implementation terms (leasing in smaller time increments) allowed a difference in kind in usage terms (on-demand server allocation over advance provisioning).
So... SaaS?
The downside to the user is that certain kinds of UI polish and more complex behaviors aren't possible. From the developer's side, there is limited ability to share code between apps, so you wind up cutting and pasting a lot, and the work is much more boring that regular dev work. Still, it's a lot less work than building up a regular CRUD app.
People get spreadsheets and word documents. I don't really know why. But they get them. But while it might seem like the step from doc/xls to nolocode is small or even non-existent, I think it actually is really big for most people who are supposed to be the audience for that (complete non developers).
I think one reason is because people don't want the responsibility to make a system that can break. I'm not too sure if this is the only or the biggest reason, but I think non-coders don't have the sense that they can build things, and even if you show them they can using some hot tool, it seems they're still scared it's going to break and it will be on them.
Maybe one reason is because if you break an Excel doc, there's usually someone around in your org who knows how to fix it. But the same can't be said for some random new hot nocode tool.
I think for specific use cases, yes. But for general, web and mobile apps, I just think it is too much complexity you can't simplify away, or at least none of the box-dragging UIs seem to have achieved that.
I agree this can be a big market, but a shift is needed to make it actually easy, not "look how easy it is" easy. Taking a different tac, maybe there's just not a market for it. The browser/HTML/JS is everywhere. It's not that hard to build a simple tool or app, but people in orgs use Excel, they don't build a simple page, even. It's not because they couldn't learn how, but somehow that just now what people in orgs do. I don't know why.
The nice thing about this model is that it implies there's a bar to be a Maker and other users can't break your work. You need to be able to reason about concepts like DRY, star schema, cardinality, etc. As a Maker in my org (15-30K people depending on where we are with the pandemic) I can support many no-code apps painlessly in addition to my primary responsibilities, whereas one CRUD app on Heroku had me feeling spread pretty thin.
For the very typical use cases there are well suited SaaS tools e.g. Shopify for shops.
I do think that the very typical cases are well covered. Sure, there's room for improvement but not sure if a real boom is around the corner but as I said I heard this so many times (and also seen so many tries) that well, at some point something may actually happen I guess.
The overarching concept of No Code/ Low Code is that "code" is the hard part of building software, ergo if we can make it so "normal" people don't have to code the problem is solved. Of course we all know that code isn't the hard part of the job- if code was the issue software developers would all be using No Code/ Low Code solutions already (as it's been pointed out, they've been around forever).
No Code/ Low Code is about selling accessibility to the wantrepreneur crowd... the same people that have an "app idea" that they want you to build. It's a huge market, but it isn't going to have an impact on software development any more than model rockets would have on the aerospace industry.
If the No Code/ Low Code tools were any good software developers would already be using them.
(When I am talking about No Code/ Low code, I am not talking about things like Webflow / Dreamweaver NOT solutions like Serverless.)
IMO stuff like serverless, OpenAPI, some of the boilerplates out there, etc are already heading in that direction. It used to take me a week to build a standard CRUD API with a handful of models, now I can do it circa 15-30min.
"Does this technology provide a means for seizing control of the existing economy?" That's the question which investors should be asking.
I think cryptocurrency is the only technology which can achieve this due to the way it transcends existing legal frameworks in terms of ownership rights; a bit like the notion of corporate personhood allowed corporations to gain an upper hand and transcend the legal frameworks of its time.
Corporate stocks were vehicles for inter-generational wealth transfer and the next vehicles will be cryptocurrencies.
It might be that This Time Is Different, but there were a lot of times in the past when "This Time Was Different" too.
I've cobbled together email sign up forms for some non-technical side projects with Google Forms, App Script and MailChimp.
I'm not a frontend developer so maybe frontend folks will tell me it's super easy, but my feeling is that there is a lot of space for low code tools that produce software that is not polished, but good enough.
I think low-code is basically just graphical programming. If you want it more powerful/universal you need to code just the same. There might just be more or less big library helpers/blocks. Maybe it's just the next step of being "high-level" and now is indeed the time for it.
Also if you manage to write something as a nice XML/yaml/json you might as well put a nice UI on it.
Integration points: Everyone sends their version to Person A, and they integrate it, filter it, postprocess it, extracts parts into another sheet, etc.
Caching: Everyone has their own version of the excel sheet in their local inbox.
Redundant: There's typically plenty of shared directories on which you can find the same excel sheet in different versions.
No, Excel is not relevant to the conversation. Excel is a standalone desktop application. No, Excel is not an example of a "no-code" webapp. Do you seriously need another adult to tell you these things?
Servicing software debt is an industry in itself.
WP lets someone with no formal CS training do wonders, and I think it's a good thing.
Cordova tries to save time and cost, but you need people with above average skill to hack things, know all the gotchas and bugs of the no-code system, all the undocumented bits, but still deal with the risks. Nobody wants to do the plumbing, nobody wants to dedicate their career to it, and so it ends up overpriced to the buyer and underpriced to the seller.
You could also end up with something like Parse, which had some of the best handovers from closed into open source, but the stigma of being dropped by a major tech company made people avoid it.
Think Prolog etc
Low code platforms, at their core, allow you to define the software with data (from predefined components usually through UI) and that's how they enable super productivity, until you hit a wall.
Data Declarative programming paradigm could potentially achieve the same levels of productivity without its limitations but it's very difficult to get the right implementation.
The most interesting development in this area I had found, is Fulcro Rapid Application Development Tools [1] - a fullstack clojure framework that's performant, declarative and extensible. RAD is still alpha but it bases on battle-proved Fulcro. Once you hit the limits of declarative RAD you can easily extend it with regular Fulcro components which are, in their essence, a cheap abstraction over React/React Native components - offering industry standard performance.
[1] https://github.com/fulcrologic/fulcro-rad
It was such a spectacular failure it's funny to see it recycled.
And for you to claim you wouldn't have to worry about maintenance is especially comical. It was extremely high maintenance and expensive, and it often broke.
Programmers don't just code a problem, they solve the problem. Teasing out the actual business logic is as much a part of the job as anything else, and one that most employees and managers are so clueless and utterly useless at that any no code solution will require a programmer to actually use it any way.
But without any of the flexibility code gives, you end up with tons of dumb, manual, expensive, hacks and a sub-standard product. Eventually the 'no-code' platform adds the ability to use some code and you end up with a really shitty software platform that delivers 80%, but the effort required to get that last 20% is far in excess of just starting again and writing it from scratch in code.
Ask any programmer who remembers workflow software their experience and you will learn from the mistakes of two decades ago.
https://thedailywtf.com/images/ads/All-programs-you-need-sma...
https://en.wikipedia.org/wiki/The_Last_One_%28software%29
That's the definition of a programmer analyst or business analyst. If the business is too cheap or clueless to create business process maps or hire someone as a process efficiency expert, then they are certainly too cheap fund properly developed software and too clueless to intelligently answer the questions that an IT resource has to use to tease out an answer.
I do agree that no-code is not going to replace the need for software professionals.
So you really haven't made the code go away; you've disguised it as something that isn't code. This actually works well for tricking people into programming when they otherwise wouldn't. And this itself is a viable strategy: consider the secretaries in the 1970s who used Lisp to "customize" Multics Emacs and were never told they were programming it. But it doesn't eliminate the complexity of coding. And most "no-code/low-code" platforms have disadvantages (vendor lock, extremely limited applicability, zero portability) that are almost guaranteed to bite you in the ass later.
This is the root cause of an endless series of failed ventures and products.
The successful no-code products are something else entirely, merely masquerading as no-code. Excel is probably the prototypical example of this: it is a reactive dataflow system, but you'll never find any dragable widgets with little input ports and output ports you can connect.
It's trivial to get started with Excel. It's trivial to connect cells and get some computations going. It's then a nice smooth transition all the way up to multidimensional analysis, and not once do you have to break out into a full programming language.
That's what people want.
What you imagine they want is something childish and useless. To be honest, it's simply arrogance. By assuming the user can't code, it's natural to assume that they think at a level of a young child and need only toys.
For fuck's sake, Excel is 64-bit and fully parallel! Did you know that? I've seen 60 year old ladies from accounting produce spreadsheets that can bring a 16-core terminal server to its knees, and that spreadsheet was doing entirely legitimate, useful, productive computation. Show me the last system you built that can efficiently utilise a modern high-core CPU for a single user!
Your market is adults. They need power tools, not toys.
Start thinking like that, and you'll have a successful product.
4GL (fourth gen languages), BPM systems, things like clipper or dbase, access, even VB is a low code kind of platform.
The new generation focuses on web tech and mobile, and they’ll do well for a while, until the landscape changes enough and they accumulate enough cruft to become unwieldy to use.
In 10 years or so we’ll be talking about something like that again.
I will be boring now and say that Cloud services will continue to expand, it's effectively a tax on doing work on the internet and start-ups love using cloud, the idea of maintaining your own servers is considered silly unless their is some particular reason to. I expect to see growth of 20% yoy in that sector for the top 3 players. Azure, GCP and AWS.
Pay a fee and you get access to a pool of desks, chairs, computers, and an IT/office staff to come to your house and maintain it and set it up.
In that vein: tooling to manage remote workers.
VR development. It really hasn’t hit mainstream yet and when it changes how we work that could be huge.
I think AI based procedural generation of games. Seeing how good gpt3 is getting this seems like it could be huge.
AI based personalized education. Have you seen how well gpt3 can explain concepts? Could something like that also evaluate your understanding, come up with custom learning plans?
Commercial agriculture is heavily dependent on, and influenced by, companies that zealously guard what they see as their intellectual property.
When I say "open agriculture," I'm referring to protecting free experimentation and open seed culture.
Open cutting exchange is a good idea actually. Trace genetics.
They’d allow for serving more direct (point to point) routes than trains as the infrastructure cost is so much lower than laying and maintaining tracks. Rail serving high volume routes still makes sense, especially while electric plans remain relatively small.
The solution is not aircraft, the solution is more and better trains. Or perhaps autonomous buses/trucks depending on how far out you look.
That leaves most of Asia outside India/China/Korea/Japan, and pretty much all of Australia/Oceania, Africa, the Middle East, and the entire New World.
I would love to see high-speed rail take over the world, but there's a lot of sparsely-populated places out there, and it makes sense to investigate technologies to support them.
In the US it'd require similar effort/cost as the Interstate Highway System did, but would probably yield equal economic benefit.
In general aircraft for passenger or cargo routes take a few routes and don't shift much.
The basic logic is two-fold:
First, it's much faster, cheaper, and less contentious to create new routes, and capacity can quickly be switched (even seasonally or in response to events) thanks to neither needing to pay for nor lay tracks.
Secondly, it shortens a lot of journeys. Even in countries with pretty extensive rail infrastructure it is common to combine two or more trips via "hub" cities in order to complete a journey. This adds time because you travel further and have to change trains and wait around. The difference could be particularly extreme given that only longer distance intercity trains tend to be high speed, and electric planes will eventually be faster even than high speed rail.
A similar change happened with long haul air travel: as twin engine planes have been able to fly further over water due to safety improvements, the traiditional hub model has lost out as many more point-to-point international routes were created.
While I think they might re-shape regional transport in some pretty important ways, electric planes won't replace high speed rail (and kerosene burning jets) on more popular routes, at least not for a while (and perhaps, debatably, never). Trains are incredibly efficient and hard to beat when you need to regularly move a lot of people, and electric aircraft, while they're about to become useful, are going to remain relatively small and range limited for quite a while.
"Somewhat"? The energy density needs to get better by a factor of like, an order of magnitude to be competitive.
And it's still going to require human pilots for at least 10 years even aside from that.
https://youtu.be/RyS92KPQnjk
Open lab equipment. Open Sensor designs.
See http://blog.fogus.me/2015/11/04/the-100101-method-my-approac... or https://www.nickbentley.games/the-100-10-1-method-for-game-d... for two descriptions of a 100:10:1 model. There are other models.
Also, we may see the rise of cryptocurrency communities which attempt to manipulate company insiders and governments to gain control of the proceeds of production to drive the value of their cryptocurrencies. Private property rights will be eroded (due to lack of enforcement and increasing systemic corruption) and so cryptocurrency, which offers a cryptographic means of ownership (which does not rely on law), will gain increasing significance.
The proceeds of production will no longer go to shareholders, they will be diverted to cryptocurrency holders. The shareholders who embrace this mindset of moving profits to cryptocurrencies will see their ownership stake of company profits increase at the expense of those who resist this shift. In the end, nobody will want to own stocks since they will no longer yield profits. Corporations will behave like non-profits; the profits will be funneled to cryptocurrencies.
All of this will be completely legal and almost everyone will support it. In 10 years, this will be completely obvious.
We are in a post-scarcity economy. Wealth creation will have little to do with productivity and everything to do with redefinition and redistribution of existing ownership rights. We will see cryptographic ownership rights surpass legal ownership rights. Once this new, highly fluid, decentralized financial infrastructure is firmly in place, the transfer of wealth will end up facilitating a new wave of massive decentralized productivity with a stronger focus on social principles.
Crypto currency is a disaster.
The benefit is that the operators of said technology can 'invest' in that bandwagon and reap piggy-back profits, while mitigating the inherent risk of actually developing those technologies or trends from scratch.