Ask HN: Do anyone get anxiety from following Hacker News?
I am 18+ year exp software professional. I see wide variety of topics in HN and I feel like I don’t know a lot of things in computer world. Off late I find it hard to follow lot of them. Crypto, ML, Data, Quantum Computing. Do anyone feel this or if you feel it how you cope up ?
59 comments
[ 2.9 ms ] story [ 127 ms ] threadIf I find something I genuinely want to learn, it’s just a matter of spending focused time learning it. I also believe it’s important to know when to cut off an interest that is no longer serving you. While discipline is important to push through discomfort and overcome adversity, it’s really silly to be wholly invested in a topic or hobby that brings you no value in life. That value can be monetary or happiness or growth. But it’s important to know that you don’t have to know everything and it’s really silly to do so. There will always be something you don’t know.
For you, maybe ask yourself why do you want to know all these things? Maybe you can identify if they serve you in a way. Of those that you believe will make your life better, spend 30 minutes every day or every other day learning from the ground up about the topic. Maybe put it into practice, or just take a course on it. Regularly reflect on if this is what you want to be doing, if your making meaningful progress towards whatever goal you have, and celebrate your successes!
> I think the key is to focus and prioritize
I think prioritizing is the key. Part of that is figuring out
What is past its prime (Cobol on punch cards), or never really had its day in the sun at all.
What is dull but you can make money with, which will probably be around in some form five years from now too (backend programming of micro-services with a popular JVM language)
What exciting things are coming into being now which might have some jobs or interesting applications now, or perhaps a lot more in a few years (training and using deep learning models).
Crypto nowadays is mostly a scam which began unwinding over a year ago. Probably best to ignore it, its day is past.
Quantum computing is interesting but it's such early days that you can put off reading about it this year if you're overwhelmed.
ML and big data have had a lot of interesting applications this year. GPT, Stable Diffusion, DALL-E. I can do some pretty cool things on my workstation with Stable Diffusion that was not able to do five months ago. If you have time to look into something new this would be it.
As I said, there's modern but dull and not much new innovation things, which there is steady work and money in like React front end programming, or JVM Microservice backend programming, or whatever. There are also things like deep learning which are going from innovation to innovation, which is exciting and may open up a whole new field (Stability AI just got $100 million in funding, although things are so new, they only started putting their hiring team together last month - they're mainly recruiting recruiters right now).
I suspect that's true of most people whether or not they believe it.
But it's only a suspicion.
Good luck.
i feel like a lot of other computing-related concepts are approaching that same level of sophistication.
i've usually been pretty good at not worrying about true understanding of various technologies, so the fact that i can't explain ai well, or that most ai is not even capable of offering an explanation of how they make their own decisions, is not so anxiety-producing to me (tho, parts of ai, and explainability, are real problems).
but i do get anxious from constantly reading headlines from various substacks and other clickbaity and attention-seeking entities about how some new tech has basically already ended life on earth so we might as well just give up.
i'm anxious enough already with real threats to the world -- nukes, global warming, social media, inequality/fascism/authoritarianism, etc.
𝐂𝐫𝐲𝐩𝐭𝐨 —- the main use case is separating fools from their money. Missing out on that one is one of the best things that can happen to you.
𝐐𝐮𝐚𝐧𝐭𝐮𝐦 𝐜𝐨𝐦𝐩𝐮𝐭𝐢𝐧𝐠 —- people starting writing papers on this around 1998 when I got my physics PhD. You could not get a postdoc in it at that time. It got fashionable maybe 10 years later. I am sure there are some people getting tenure in it right now but in 10 years it might be remembered as just another field that ran its course and had better days and it may get a reputation, just as chaos and fractals got in the 1990s as something you just could not make a career in. Sure there will be ‘quantum supremacy’ for particular problems (maybe even that box that is the McGuffin in Sneakers) but you are not going to be using a quantum web server to host your content.
𝐃𝐚𝐭𝐚 —- I hear ‘data is the new oil’ and I think of an oil tanker spill like the Exxon Valdez or maybe like Spider-Man, ‘with big data becomes big responsibility’. I think how TikTok’s recommendation algorithm is so good because they collect better data: they give you a very limited set of choices so the data collect is meaningful. When YouTube shows you 30 links on the other hand you can come to no conclusion that the viewer didn’t like those things.
Data is a big part of everything we do in computing. In plain ordinary boring applications development the most important thing is getting your data structures straight. There is a whole art of online transaction processing and the arts of doing analytics after the fact or in real time.
𝐌𝐋 —- data is the currency of ML. If you have a real problem to solve and real test and training data you can do ok with basic algorithms out of scikit-learn. If you don’t have appropriate data you can dream about the latest algorithms and accomplish nothing. You need to know the basics of linear algebra and statistics: the courses you take as a math major are great.
Like most other things there is a part of the field which is almost eternal (I think about things Yan LeCun wrote about 20 years ago almost every day and think ‘Neural Networks: Tricks of the Trade’ can turn a failing project into one that succeeds even though it was written before deep networks.) and other parts that are ephemeral. For work I did a lot with LSTM networks for text, then it was CNN networks, those have been forgotten almost because transformers can do things that would be difficult or impossible without attention. Personally I am really interested in those radiance field algorithms for scene synthesis, I haven’t done a project yet but it was clear early on that the algorithms would get much more efficient and that has been happening the whole time I’ve been watching from the sidelines.
There are ‘data scientists’ who do quantitative marketing with a lot of smak and pow and get paid really well because they really can increase revenue 850% in two years. They don’t necessarily know a lot about algorithms. If you were working in the field as a scientist you would be one of the people who writes one of the 20 papers on radiance fields a week that would never get a second vote on HN, but as a practitioner you can master the basics and solve real problems for fun and profit.
I feel you first need to learn how to be competent, then you need to learn how to be incompetent. By that I mean recognize your areas of deficiency, know how and when to ask for external support without being ashamed about it (being competent _somewhere_ helps with that). I don't know if this is meta-cognition, wisdom, or something else.
While you're still struggling on the first step it's easy to get tripped on the second one.
My recommendation to younger people starting out is to aim to become a "T Shaped Person"[0], or even better "F", or "π" shaped. The idea being to have a very wide range of shallow knowledge, enough to communicate, understand, spot the BS, and make a start. But have a very deep area of expertise in one, or ideally a few, core areas relevant to your chosen field.
Personally I find HN an invaluable resource to ensure that I have a very board, but shallow(ish), understanding of a large number of interesting things. I've been here for 14 years and it's probably been one of the most valuable experience I have had.
0: https://en.m.wikipedia.org/wiki/T-shaped_skills
Definitely. You spend a couple of years learning new tech and methodologies and then the industry moves on and you’re behind on something different.
Ai is going to rock our world. I personally don’t know what to grab hold of to keep from getting tossed overboard during the coming storm.
Great time to start an ai company though, if you can get financing. If you’d like to fund me, I have video game related ideas I’d pursue if I could ^^
So following the HN community too closely can stir feelings of insecurity and inferiority and the imposter syndrome. And yeah, I feel it sometimes, too. :-)
I would dabble in many other domains like AI, blockchain, other languages, etc. to test out their effectiveness, but the best advice I have is to focus on your problem first and then find a technical solution that solves that problem. If you start with the technology first and then go looking for a problem, it’s not as effective IMO.
I cope by reminding myself that I shouldn't compare my personal knowledge with the combined knowledge of a community - multiple people here are knowledgeable about and post about different areas
The best way of coping is a JOMO (joy of missing out) and moderation. You don't need to know everything and you surely don't need to spend all your time trying to. It's okay to not know. That's the beauty of life.
I've been in this space since 1995, and I've learned that most of the things that pass by simply don't interest me. I don't really keep up with JS/CSS/HTML stuff any more, since it's a saturated space in a work sense. I focus on the back end exclusively, where things move in much slower, provable steps.
some of these things, are things you wont ever need to know, but observation is enrichment.
part of a critical knowledge base is allocation of effort toward things you need to know for progress, and indexing those things you dont need to know right now.
I'm here for the smart(er) people with experience in different areas. The thing that makes you anxious is the reason I enjoy being here.
Crypto seems to have outlived its utility as a way to get rich quick. ML is still just a technique to average out giant datasets to find/generate things based on trends (these things work on training data after all and not from a true blank slate). Quantum computing is a long way off and the most practical way it’ll affect most people here is making our encryption keys too simple and some new quantum resistant version will be needed.
Just watch chatGPT - in 6 months, it’ll feel like a distant memory here! And people will be raving about some new JS/wasm thing instead
> Also that the HN bubble will always make readers feel behind when in reality 90% of the industry hasn’t even begun to catch-up. I felt behind with K8S in 2016, for example, not realizing we were just seeing the wave form when I felt like it was cresting. [0]
0 https://news.ycombinator.com/item?id=23448835
I wish I could filter these out and just stick to the technology.
You can't know everything and you can't keep up with this firehose of new knowledge that is published every day.
Most of that is not relevant to your field anyway - whatever it might be, so you don't actually have to know any of it.
I mean, you're free to take it all in, but don't expect anyone to quiz you on it in a professional environment.
I could imagine fears about job security, maybe some sense of falling behind? A simple fear missing of missing out on all the fun in specific sub-fields?
We have such a large field in general, I don’t think anyone can know it all, but that can be something to celebrate too. Never a dull moment :)
Not really surprising considering the vast majority of people here seem to have probably not received higher education (university/college) in regards to STEM. I didn’t think this was the case before I joined though.
You got any critical thoughts or evidence for that conclusion?
There are people here with a compassionate, humane understanding of technology, who see the dangers alongside benefits and get the complex social implications. Amongst the general population, and even within academia, I often feel isolated, but people here give me enormous hope and courage that I am not alone as a critical thinker. Their good humour lifts my spirits.
A smaller group are callous, parochial, selfish but dangerously intelligent people with a Dr. Strangelove style fetish for Brave New Worlds, Singularities and Technological Master-Races. They believe they are a "majority", that others should get out of their way, and they hide behind the banner of "inevitable progress". They would steamroller over the weaker, poorer and less intelligent margins of humanity, burn freedom and sell their souls to make a buck. They are also humourless and prickly in defence of technological golden-calfs. En masse they constitute a blindly-progressive mob-rule centred around technology that terrifies me as much as any far left or right wing political or religious extremists.
On balance, I think the former, reflective group have the upper hand. All kinds of opinions are fairly represented here in HN. That co-existence and mutual recognition gives me significant hope.