Marc Kelechava

> The real story of this era will be who manages to avoid harming themselves in their AI psychosis. this checks out and i see this in a few programmers and non-programmers i have met over the years. programmers feel they can do everything themelves. non-programmers suddenly feel they can program production code. LLM psychosis induces hybris, should be met with humility, IMO. > Without fully endorsing all their ideas, I’m now in the LeCun/Marcus camp on LLMs. I don’t think models like this will ever be able to program, I think the process matters. I think that deep learning is still the solution, but real programming agents will need world models, not some RLVR shit that comments out the failing test and tells you all the tests are now passing. i dont know enough about world models (but i will study up now!) but i think the part about "LLMs get 80% right and then screw up the polish" is consistent with my experience and i wonder if it's a fundamental limit i don't know if i resonate so much with george's writings because we have the same name or believe similar things, or because i used limera1n as a kid but the last few writings hit the spot for me geohot.github.io/blog/jekyll/up…

You can tell AI is a net good for society because mark zuckerberg is bad at making it

You might believe you should spend less time thinking about code because of AI. I strongly disagree! We’re watching this play out live where tons of AI generated code becomes a liability. At the end of the day, an engineer needs to be responsible / on call for code that gets shipped to production. If you don’t understand the system you’re trying to debug, you’re probably going to have a bad time. Yes, AI can help with all of this, if you set up the proper systems. You can have agents triage prod logs, look at errors, etc. You can speed up parts of the investigation, but an engineer needs to make the call. There might be serious customer or financial implications from that change. I expect the trend continue for trimming dependencies, vendoring code so you can modify it directly, preferring simpler systems with fewer abstractions, and spending waaaay more time thinking about system design and code maintenance. I’ve said this before, but it’s a great time to get familiar with CS fundamentals and some of the history behind what great software looks like. Many parts will be different in the coming years as AI progresses, but also a lot more than people realize will stay the same.

there is no better time in tech than now to be a jack of all trades, master of a few. just make sure to keep adding to the few year over year, such that the cumulative breadth of expertise you collect becomes an increasingly rare combo. remember, if you're top 10% in 3 different areas, that already makes you top 0.1%. keep switching it up until you get to "your best", and then switch it up again (great for a particular flavor of people who don't enjoy resting on laurels, maybe not so great for others). question all institutional value and pedigrees, all traditional career paths or corporate ladders: the college industrial complex is getting shaken up, alongside a disappearing managerial class, so if you're pursuing either make sure you are fully internally aligned with why. social/political capital in a particular institution can feel incredible, but if you're spending all your energy on complex political people games, you're not a technologist anymore, you're an unelected politician. if you're ok with that, then all's well. critical thinking is more important than ever: take nothing at face-value, question everything and everyone. the equivalent of ai slop can be found in humans operating under misaligned incentives and interests. the sooner you're clued into disambiguating the talkers/larpers from the doers, the better off you'll be figuring out where and who to invest your time in. the anxiety of job displacement is very real, since a surprising amount of white collar work/prestige is built on a performative house of cards, significantly lacking in correlation with technical breadth, depth, and skill. as long as you keep learning, keep building, keep producing receipts, you will be fine. if all that sounds ok to you, welcome to the world of technology! it's truly one of the few places you can experience child-like wonder every few years, and be constantly humbled & excited by new adventures, as scary as they may seem at first. don't give up, drink your water, get your sunlight, and take breaks as needed. tech careers are notoriously nonlinear, so you might as well embrace it and enjoy the ride!

@RaeAlisa_ @lucent_ai codex

to celebrate 3 months since lauching @lucent_ai, we're giving away 5 Codex Pro / Claude Max plans 🎁 to enter, like this post + comment which one you'd pick (codex vs claude) winners will be selected from comments in 5 days 🫶

how do u fix openclaw internal reasoning leaking

sf dudes be like "taste is all that matters" while having one patagonia fleece and a datadog summit 2022 t-shirt to their name

I quite like the new DeepSeek-OCR paper. It's a good OCR model (maybe a bit worse than dots), and yes data collection etc., but anyway it doesn't matter. The more interesting part for me (esp as a computer vision at heart who is temporarily masquerading as a natural language person) is whether pixels are better inputs to LLMs than text. Whether text tokens are wasteful and just terrible, at the input. Maybe it makes more sense that all inputs to LLMs should only ever be images. Even if you happen to have pure text input, maybe you'd prefer to render it and then feed that in: - more information compression (see paper) => shorter context windows, more efficiency - significantly more general information stream => not just text, but e.g. bold text, colored text, arbitrary images. - input can now be processed with bidirectional attention easily and as default, not autoregressive attention - a lot more powerful. - delete the tokenizer (at the input)!! I already ranted about how much I dislike the tokenizer. Tokenizers are ugly, separate, not end-to-end stage. It "imports" all the ugliness of Unicode, byte encodings, it inherits a lot of historical baggage, security/jailbreak risk (e.g. continuation bytes). It makes two characters that look identical to the eye look as two completely different tokens internally in the network. A smiling emoji looks like a weird token, not an... actual smiling face, pixels and all, and all the transfer learning that brings along. The tokenizer must go. OCR is just one of many useful vision -> text tasks. And text -> text tasks can be made to be vision ->text tasks. Not vice versa. So many the User message is images, but the decoder (the Assistant response) remains text. It's a lot less obvious how to output pixels realistically... or if you'd want to. Now I have to also fight the urge to side quest an image-input-only version of nanochat...

Accidentally said "random" instead of "stochastic" and they kicked me out of SF

one of my favorite parts of AI is never having to write matplotlib again

codex just made a python script to edit one of my tsx files, what are we doing man

americans saw water freeze at 0°C and said “let’s make that 32”

👀?

What do you think?

Some people are so dumb that no AI model could ever replicate them.

if you are so smart why are you not only doing things you enjoy

Stocks trade based on their fundamentals

yo bro I have a cool idea for an app ill give you 50% equity if you build everything and do all the work