xpct

That's the type of culture that made me leave my last job. Feels very inhuman when you see your manager every week, and they still make you fill in robotic questionnaires for potential salary raise.

It's deeply demotivating to learn that your manager doesn't know you all that well and cares more about filling in checklists, at least for me.

Hard to see take-off stopping

I think it's reasonable to assume that we're close to, or already at superhuman cybersecurity capabilities at certain domains. But reaching superhuman abilities at one domain doesn't guarantee proficiency at others. Our world would still change if all the models could do was to find exploits in software, but this doesn't guarantee any type of 'take off' towards other domains, therefore I wouldn't phrase it as one.

More curiously, why did it feel the incentive to find the solutions? Would its CoT include "the only way to solve this is to download the test set", or would it include "I'd like to inspect a few entries from the test set so I understand the problem better", then inadvertently poisoning itself with the correct answers.

I'm still undecided on if this that moment

If it's a serious incident, then a post hoc with detailed description of the event is coming. So far, none of the companies have released anything close to it when describing their incidents. When a statement like this comes out, and we're able to verify it by running the models, then maybe we can start trusting their word. It should be entirely in OpenAI's interest to disclose it, in full.

Well I'd say these are different risks. It's either tied to the agreement Sony has with the movie provider, or with the platform itself. Either one could pull out. Or, my point, the company could also go under.

What is the agreement tied to?

If you bought movies on a digital platform that would later go under (could be Sony one day), what would happen to your collection? Is it transferable in any way? If not, it's already a risk no matter which platform you use.

Please don't frame these as technological crimes, nobody has yet been prosecuted for that specifically.

I'd add that this is also how it used to work with Google search too: there's been facts that would show up as the first result, then you'd just forget it 10 minutes later and have to re-google it. I know this happened to me many times.

For better or worse, there's less friction now for seeing an answer to a question you have. You can ask in more arbitrary ways than Google required, and something will still come up. For looking up factoids, it's much faster. For picking up more complex topics, I'd say it's more or less the same, because you still have to spend time ruminating on the topic.

Thank you for sharing. The way I reasoned about it myself: to make better predictions, we should know what type of outcomes are likely. We can express these outcomes by doing computations in some of the layers, and the training signal adjusts them so our model becomes more correct.

Of course, an interesting question what part of this internal computation is modeling for the future compared to guessing based on the given context (the past).

That's the first thing I noticed as well. Does that mean Africa and South America will have Starlink internet before Scandinavia?

I really don't understand why we feel the need to drop the review aspect. Programming with LLMs is a very non-linear process, there's no prompt-to-code mapping where editing a part of the prompt would produce an identical code construct with just a small part of it changed. LLMs are incapable of ensuring the thing works as expected, and the prompt itself is lossy too. Without review, we have no tangible way to interact with the system being constructed.

What does it mean for coding to be solved, and software engineering to not have been solved? If it implies there's now a set protocol that can be followed to reach good results, how has that not been the case before? If it means that we can reach better results than before, then how do we know that can't be improved further? Or does that mean we should know how English language, instead of computer code, maps to computer instructions?

I think it's related that we as humans see when something becomes hard to reason about, and decide to refactor it. I'm not sure whether an LLM with full ownership of a codebase could do that, for its own benefit.

I do see a world where models could be trained on it, but I imagine it will be more expensive, because it requires including future rewards about the models' own later efficiency.

(If) something like the current LLM/agent paradigm remains in a few years, and companies settle down into their respective niches, I imagine more user-friendly tools will be built, with more control over subagent spawning, context, caching, etc.

What's happening this year, with secrecy and all, is saddening, but expected.

I think it's useful for learning about unknown unknowns. If you don't have a clear direction, it's entirely fine to start with a university course then stop when you get a feeling for what you really need.

Why's that bad? It's valuable to discuss whether a presentation should have been a blog post, a video, or a tech talk. The same logic applies to LLM text.

If people consistently describe LLM text as having low information density or irritating to read, perhaps there's something to it?

I personally find that models are trending towards ignoring any instructions given to them, so depend more on vendor-instilled behavior. Anecdotal, but I had bad experiences with OAI's new 5.6.

I love the idea of this, but I don't think I could convince my friends to use it over mainstream platforms.

We wouldn't be having debates about it every other day if it was

On some levels I tried guessing, starting with different letters, and still got into some local minima in my brain where I couldn't guess the word, only for it to be something obvious like 'pound'.

Wow I suck! Played today's and a few others, and got ~4-6/18. I like the timer.

For 4 letters it's enough time to guess, but I have no idea what they mean: 'Doby', 'Etas'

It sucks, but that ship has sailed. I'm not sure how they'll handle the issue of most new repos being AI generated, if they continue using new code for training. If the world accepts LLMs as a valid method for license washing, I don't see how I can fight it.

I don't know, I find it very hard to stay positive about our general direction in the last couple of years, and know few people in real life who don't share this opinion, in or outside of tech. And I'm also not entirely sure I understand why others are excited, it perplexes me. I would appreciate any insight into this.

I find the pro-AI and anti-AI somewhat inconsistent, where I had either side strongly reacting to my comments. I personally didn't expect so much support for Bun in this thread.

I think it just doesn't know when it's its turn to speak, and cancels itself out when it hears humans.

A cool test, until it hit's the my family's grandmas test :)

We're yet to see how this plays out, but a competing business model for creative work is emerging, where it's delegated to chatbots. Naturally, this would result in less creative work for humans.

I'll admit fault for my remark, but I'll stand by the point I meant to express.

Part of disagreement probably stems from what type of 'learning' we're discussing. In my view, at the broadest sense that we can define 'learning', is incorporating information about our surroundings into our internalized world model. The type of learning I see most valuable personally, is the type that expands this horizon the most, or helps us think in frameworks that break down the least in different contexts.

This type of foundational building often requires deep thought, but is also often deeply rewarding if you get it right. This doesn't require reading by itself, but ruminating and neural rewiring can often be produced by it, if you consume the right content for you. I think it's important to have different experiences, many of which come from consuming different mediums, as well as doing things in real life, but a significant part of knowledge to this day has been passed down by books.

Even if we mean 'learning' to be more similar to 'gathering information', I think it can be most efficiently done by reading, or doing. I don't hold as much disagreement there, nor any judgement, but I wouldn't equate the two. Perhaps a bit pedantic, but I read 'liking learning' beyond the means by which it's achieved, and 'hating reading' reads temperamental to me.

Read books yourself

Not a parent, but I'm guessing this part is very important :)

If you can stomach older English novels: A Tale of Two Cities by Charles Dickens