HN user

leshow

2,116 karma
Posts1
Comments1,119
View on HN

If you want my opinion, I think LLMs can be pretty good at generating simple code for things you can find on stackoverflow and require minor adjustments. Even then, if you don't really understand the code you can have major issues.

Your site is case in point of why LLMs demo well but kind of fall apart in the real world. It's pretty good at fitting lego blocks together based on a ton of work other people have put into React and node or the SSE library you used, etc. But that's not what Karpathy is saying, he's saying "the hottest programming language is english".

That's bonkers. In my experience it can actually slow you down as much as speed you up, and when you try to do more complicated things it falls apart.

It's not the same, LLM's are qualitatively different due to the stochastic and non-reproducible nature of their output. From the LLM's point of view, non-functional or incorrect code is exactly the same as correct code because it doesn't understand anything that it's generating. When a human does it, you can say they did a bad or good job, but there is a thought process and actual "intelligence" and reasoning that went into the decisions.

I think this insight was really the thing that made me understand the limitations of LLMs a lot better. Some people say when it produces things that are incorrect or fabricated it is "hallucinating", but the truth is that everything it produces is a hallucination, and the fact it's sometimes correct is incidental.

Async DNS 7 months ago

I use hickory a lot and have contributed to it. It does have a pretty robust async DNS implementation, and its helpfully split into multiple different crates so you can pick your entry point into the stack. For instance, it offers a recursive resolver, but you can also just import the protocol library and build your own with tokio.

I love Rust and async Rust, but it's not true that there aren't annoying things to deal with. Anyone who's written async Rust enough has run into cancel-safety issues, the lack of async Drop and the interaction of async and traits. It's still very good, but there are some issues that don't feel very rust-y.

That would be the case under market conditions where buyers are making rational decisions with perfect knowledge based on all available choices. Does that sound like the system we have? To me, reality seems more like a small set of oligopolies or effective monopolies, byzantine ownership structures and a pursuit of short term profits pushing future costs elsewhere as externalities.

AI is different 11 months ago

I think you're confusing publicly available shares with actual ownership by the public in a way that's no consistent with the way any socialist would think about ownership. Stock may be available to anyone in the sense that anyone could buy shares, but actual ownership and decision-making is highly concentrated. Real public ownership would mean something quite different.

AI is different 11 months ago

It's the same here. Calling what the west has a "free-market capitalist" system is also a lie. At every level there is massive state intervention. Most discoveries come from publicly funded work going on at research universities or from billions pushed into the defense sector that has developed all the technology we use today from computers to the internet to all the technology in your phone. That's no more a free-market system than China is "communist" either.

I think the reality is just that governments use words and have an official ideology, but you have to ignore that and analyze their actions if you want to understand how they behave.

AI is different 11 months ago

In your world where jobs become "optional" because a private company has decided to fire half their workforce, and the state also does not provide some kind of support, what do all the "optional" people do?

AI is different 11 months ago

UBI is not a good solution because you still have to provision everything on the market, so it's a subsidy to private companies that sell the necessities of life on the market. If we're dreaming up solutions to problems, much better would be to remove the essentials from the market and provide them to everyone universally. Non-market housing, healthcare, education all provided to every citizen by virtue of being a human.

I kinda feel this way too. Reading some of the blog posts by AI "luminaries" I'm struck by how Stalinist they sound. They hold out some utopia that exists in their minds, and they are ready to feed people into the meat grinder to try and make it a reality. Stalin said that this generation would suffer so that the next lived in utopia, and that's kind of the same pitch they are making.

I think if we actually cared about making a better world, you'd take steps where each successive step is a positive one. Free healthcare, clean energy investments, etc..

I don't think that's a fair comparison. Type systems don't produce probabilistic output. Their entire purpose is to reduce the scope of possible errors you can write. They kind of did change the world, didn't they? I mean, not everyone is writing Haskell but Rust exists and it's doing pretty well. There was also not really a case to be made where type systems made software in general _worse_. But you could definitely make the case that LLM's might make software worse.

I think you're missing the broader context. There is a lot of people very invested in the maximalist outcome which does create pressure for people to be boosters. You don't need a digital token for that to happen. There's a social media aspect as well that creates a feedback loop about claims.

We're in a hype cycle, and it means we should be extra critical when evaluating the tech so we don't get taken in by exaggerated claims.

I have a degree in software engineering and I'm still critical if its inclusion as an engineering discipline, just given the level of rigour that's applied to typical software development.

When it comes to "prompt engineering", the argument is even less compelling. Its like saying typing in a search query is engineering.

The state keeps the billionaires afloat though. Tesla, SpaceX, all of Musk's companies wouldn't exist without government welfare. Other big monopolies benefit from government protection and subsidy.

Billionaires are not a fact of nature, they exist because the rules we have created allows them to exist. Different rules could make them not exist, and society would benefit hugely.

Regardless of your feelings about Elon, what would the strategic implications for the US vs Russia and China be minus SpaceX?

Weird to think of everything in the view of great power competition. Like, you can't have healthcare cause spaceX needs to beat China? There are many options.