A lot of devs are micromanagers who think they’re smarter than everyone else. You’ve probably at least met a couple if you’ve had a long career. They typically leave lots of nits on code that is mostly meaningless noise or personal preference.
These types of people will get the least out of AI as their ego prevents them from letting it outshine themselves.
I am sure people could find code such that an LLM can optimize it 10x fold. I have seen many types of code in my life (well written, horribly written, fast, slow, etc.).
I do feel there is a limit to LLM-s today. I did not try, but I doubt it will work if I would tell it "make me a web browser that is bug free, perfectly secure, works as efficient as possible on my architecture and has best UX for me personally".
Knowing where that limit is, is as hard as it always was knowing how fast a team of engineers was going to do a project.
I felt there were limits 4 years ago. I couldn't get even an 8k context window.
It's starting to feel like the important gaps are the only ones left that need to be closed before there aren't any left. It also feels like next year they will start meaningfully closing.
> I did not try, but I doubt it will work if I would tell it "make me a web browser that is bug free, perfectly secure, works as efficient as possible on my architecture and has best UX for me personally".
I do feel there is a limit to human-s today. I did not try, but I doubt it would work if I asked literally any programmer I know "make me a web browser that is bug free, perfectly secure, works as efficient as possible on my architecture and has best UX for me personally".
Hell, I'm willing to bet they'd fail at this task even with an unconstrained snacks and kombucha budget. Humans have a long way to go. My job is safe.
This doesnt even apply to his issue. What does PyTorch code have to do with complex concurrency or complex systems in general.
Theres lots of bad Pytorch code, just go ask george hotz. This isnt the magic you think it is. And nobody wants to hire the guy that just points llms at things and says make it faster. Things have value because talented humans make them. Its why a luxury coat is worth more than the linens that make it, or one from walmart made by a machine. This will 1000% apply to programmers. I think OP will be fine.
Not every optimisation is the same. LLMs can optimize code that is behind known practices, but won't usually invent new data structures or algorithms to solve a unique problem.
While it's unlikely for LLMs to implement patterns not present in their training, current training techniques absolutely can produce practices and data structures and algorithms that humans have never done before. Modern LLMs are not just trained on text. They are trained on computer use, including - at a high level - solving programming problems with code. They are not just rewarded for producing the most likely next token. They are rewarded for producing tokens that solve problems. Patterns and techniques can emerge from this that are not present in any human behavior.
They are rewarded for code compiling, thats about it. Thats why the code is terrible. Ive yet to look at a project made by an llm whos user swore this "frontier model is the one", where the code wasnt utter garbage. The heavy llm users dont care as long as it compiles.
The fact that you're pointing it to PyTorch code in the first place is being specific.
Pointing a frontier model to a 10 or 100MLoC codebase and saying "make this code fast, make no mistakes" doesn't work. As an experiment I recently tried this with a relative small (500KLoC) codebase and it got stuck on believing that the primary cause of slowdown was the database not using a connection pool. (Which was completely irrelevant for this specific code.)
I’m happy he’s dumb. If he were smart, he’d be way better at corruption. His lowbrow corruption of scam meme coins and fleecing tech CEOs for a couple million is pretty benign versus what an actually smart person could do.
No. His real innovation was shamelessness. A lot of high society functions on these "gentlemen agreements" where you can theoretically get away with murder but it'd be rather uncouth of you to do so. He doesn't care and it's led him very very far in life.
Yeah I wouldn’t be so happy about that. He’s that dumb because he was chosen for the dumbness and plausible deniability
There are much more incidious actors behind the scenes doing corruption at levels we can’t fathom whole Donald is the blunt force tool absorbing all the attention and negative press with his goofy corruption
I definitely can't use the word "benign" given the immense damage to our culture, economy, and national standing he's done, but directionally agreed. We're fortunate he's an actual stupid person.
When has any Facebook executive ever gotten a prison sentence for their horrendous doings? They just pirated all of Anna's Archive to feed into their AI.
Our legal system has been turned into a tool to punish the poor and enrich the wealthy.
Meta, Google and Apple are all currently claiming immunity in court for the explicitly illegal stuff they distribute and profit from - from scam ads and apps to illegal gambling apps.
I’m up for anyone and everyone pointing it out until it’s fixed. Which is hopefully one step on the path to action, which is on the path to change, and eventually resolution. Justice and fairness will be an ever present battle or at least vigilance.
You mean one shotting slop that barely works and lacks mature software stability with decades of fixes/features and improvements? No serious professional would ever use something like that. Like I’m the first to join the party if there is any hate to put on Adobe, but their closed source platform isn’t going anywhere. Only a paradigma shift could
I don't get how people are still denying what's happening in front their eyes to this day. "Slop that barely works" isn't holding up as much, model output even in the hands of novices is drifting away from the "slop" moniker quickly. I downloaded .2 a day ago and none of the preferences worked, .3 just now fixed more than half. 3 years ago we had a mangled mess of Will Smith eating spaghetti and now audiovisual content is more or less solved... age old adapt or get left behind.
We're one major LLM release away from being able to spit out _the whole adobe suit_ in a single prompt, while being safer, faster, and more featureful. I wonder at what point will tech companies stop interrogating candidates on useless Leetcode problems.
Ok. Given this hypothesis, why is the software they release generally considered crappy by competitor standards, benchmarks, and open source standards?
Claude Code is an awful codebase, has leaked its own source code multiple times, and scores the worst on number of tokens burned vs pass rate percentages.
Probably because bad code that you create initially without thinking that it's a core piece of your stack becomes depended on for its crappy behavior, and then you can't change much without breaking workflows.
I seem to recall Fred Brooks talking about that experience with OS/360 JCL (maybe just straight up in The Mythical Man Month?).
If agents really are superpowerful at programming tasks why not just have it rewrite the tool that the majority of your customers use and have it recreate the bugs? I mean presumably its the primary force behind the current version so what's the major cost there?
I imagine it comes down to economics.. there isn't much upside to fixing the last 20% of issues that the dumber faster models are missing.
The cost to serve, latency profile ,and internal demand for a maximal intelligence model would probably keep it pointed at harder and more valuable problems most of the time.
I mean, on the one hand the tool may be vibecoded crap - I don't know, haven't checked, taking it at your word.
On the other hand, Opus 5.5 cracked zero-shotting proper LCARS interfaces that near-perfectly adhere to the franchise "design language" even in tiny details, while simultaneously being 100% functional following my admonitions about Airbus cockpit design rules and nuclear reactor control room standards.
So yeah, why wouldn't I use it? It works spectacularly well.
Absolutely spot on. Same thing goes on with interviewers trying to one up, trick, and feel superior to the candidate they’re interviewing instead of assessing capability/fit.
When Astra and Fable stop making absolutely brain dead mistakes on simple tasks and not a moment sooner because I’m grounded in reality. Are you aware tech bros have been saying this same thing about every AI advancement since the 1950s? The reason being every AI researcher deep down wants to believe they’re God. This happens to also be why they’re some of the most insufferable people.
That's a false sense of security. Something could have a serious issue, but provide overwhelmingly more benefit (cars, say), and thus people would use them anyway. If people did not use cars, then traffic fatalities would not be a concern.
People are using AI.
Your argument is that AI cannot be a serious concern because it has major known limitations. But most of us are using it despite the major known limitations. So the existence of the known limitations is irrelevant to determining what level of harms are possible -- unless the known limitations are directly limiting the harms. In practice, it seems like most current harms, and many of the (not stupid) feared harms, are completely possible despite the known limitations. Sometimes even exacerbated by them. ("Sorry, you told me to never stop your insulin pump, but I stopped your insulin pump while diagnosing a test failure. I will record in my memory file that I must always... hello? user?")
The (not stupid) parenthetical might be too weaselly for you, but that's how I see the "AI is going to rise up and intentionally subjugate or exterminate humanity" brand of doomerism. While I don't think that's strictly impossible, there are so many other ways harms that are much more imminent and much harder to defend against, that it seems foolish (to me!) to worry about those.
If you can stop a problem by pulling out the plug, you don't have a real problem.
If you don't want to pull out the plug, you do have a real problem. (Don't want, or can't afford to, or can't be bothered to so extremely that you never will, or whatever.)
> Are you aware tech bros have been saying this same thing about every AI advancement since the 1950s?
Whatever AI people have been saying for the last 70 years, AI capabilities have improved stupendously in the last 10 years and there is no sign of the rate of improvement slowing down. I would like to deny it, but I can't.
> I’m also very concerned about other AI-related issues: cognitive surrender, AI-induced psychosis
How's he so sure he's not already in both of these groups? His post about the AGI boogeyman sounds like it to me. Why is he bothering working on both of his projects if AGI is right around the corner anyways?
These types of people will get the least out of AI as their ego prevents them from letting it outshine themselves.
reply