HN user

bykhun

58 karma

https://me.bykhun.com @caffeinum on other platforms

[ my public key: https://keybase.io/caffeinum; my proof: https://keybase.io/caffeinum/sigs/xhI_n7dLyJd2gGJg6a1lF9TkIxgSg4mA6AK-UxwOc_k ]

wonderfuldev_jjqfqwo90md6qu7nazyswpl7

Posts4
Comments28
View on HN
Arm AGI CPU 4 months ago

Certainly, they wouldn't expect that an AI able to score 150 on an IQ test is unable to play a casual game of chess because it isn't coherent enough to play without making illegal moves.

To be fair, I am pretty sure Claude Code will download and run stockfish, if you task it to play chess with you. It's not like a human who read 100 books about chess, but never played, would be able to play well with their eyes closed, and someone whispering board position into their ear

My project is a complicated mix of account abstraction, telegram mini apps and about five obscure cryptocurrency SDKs integrated throughout the app. Yes, Cursor frequently hallucinates but it’s pretty easy to steer it. @Docs solves it.

Another piece of magic sauce: waste the context. Spend the tokens as if you’re partying last time in your life. Throw anything remotely relevant at it.

I haven’t yet found piece of context that made it perform worse. Most of the time the opposite happens - it finds stuff that I haven’t noticed in the big chunks of data I threw at it (error logs, documentation, stack traces)

I remember when I was just learning programming I had a mental block around nested for-loops. I couldn’t trace more than 1000 operations in my head, so I thought that computer wouldn’t be able to as well. To my surprise a computer can easily handle a million operations in a split second.

With AI is a similar feeling, I feel constrained by the amount of my own “RAM”, but once I let go, it almost always surprises me.

I'd also argue that the biggest threat from ASI is what I've heard Roman Yampolskiy label as "ikigai-doom"; that AI could become so much better than humans at all the things humans do, that even in the best case humanity is left with no purpose

I bet status games will stay. Robots may be sexual partners, CEOs and therapists, but they'd never take on status roles in our society – only utility roles.

Same as we do Olympics, even though machines are much better at throwing and lifting than us – we do it to win approval of others

I was quite willing to work for a lot less than many folks much less qualified, on account of having my retirement set, and was looking for interesting and engaging work, more than money; but it never got that far.

Did you think to work with blockchain? It has a lot of engaging problems, and you can start even without getting hired. It ranges from helping out DAOs in open-source and getting paid in tokens; to starting off your own small project and building a community.

As for the jobs, the compensations has been through the roof lately – the field doesn't get new developers nearly at the rate needed

I see what you're aiming at, but technically it's not true. It can be similar though.

The point about how we can only infer their existence from a mass of data is similar to how we talk about deep learning. The pretrained neural net works, but we can't pinpoint why, we can only test it on various data and receive results.

There is no "dog that didn't bark", because we clearly can see the graphs and the correlations.

The way we detect gravitational waves is not "the photon haven't arrived", but "the photon has arrived a few nanoseconds earlier than we expected it". (But multiplied by a few thousands of reflections and detected only by interference pattern, not the photon arrival itself).

So, in a way, it's similar to deviations from paths by which we presume dark matter. But also, it's not that: we are almost certain here that these exact deviations are caused by the gravitational waves, because calculations and because multiple unconnected detectors, and because the form of the graph matches.

When I saw a title, I thought that was a story of a learn-js-in-10-minutes, haha.

I have a degree in CS, but the more I read, the more I could relate. I believe the good formal education is very valuable for a software developer, but it's true that most things I learned were from internet and self-education on examples.

The degree is a great thing to have, but when you're native speaker in Javascript, it's even better.

Still, these things are not replaceable: I believe both are important.

Now, I'd like to be a party-breaker for a minute and counter-argument the idea that the degree is redundant in place of a developer experience: it gives you structure and a framework for a problem-solving. I am able to mock-up and prototype quickly, and to build systems because I have received formal 6-year education in maths, physics and CS. I haven't had a Javascript classes. I learned it on examples, starting from HTML forms in 2007. But the things I'm able to do now in React would be impossible without the background in GTR, quantum physics and calculus.

Yes, you would be a great developer without a degree. But you would be even greater one with it.

It is very good! Tried it yesterday, and opened again today.

Now I want is as a soundtrack for my life :D You should totally publish a Desktop app which loads on system boot!

Also, did you think of partnering with one of these relaxation or meditation apps?

I have not researched the market, but I have thought about it quite a bit. I think there is one main advantage that is trust.

Say we want to build a decentralized Google so that noone could own people's personal data and yet we train data-hungry AI for search, face recognition, ad recommendation etc.

I have thought about this idea quite a bit, but the current level of hardware (and software) is not enough yet.