Live data from Hacker News

Reports of code's death are greatly exaggerated

stevekrouse.com

411–420 of 486 posts

Re: Reports of code's death are greatly exaggerated

#411

Earlier quoted context omitted.

You know where LLMs boost me the most? When I need to integrate a bunch of systems together, each with their own sets of documentation. Instead of spending hours getting two or three systems to integrate with mine with the proper OAuth scopes or SAML and so on, an LLM can get me working integrations in a short time. None of that is ever going to be innovative; it's purely an exercise in perseverance as an engineer to…

Itops related work in general is so suitable for ai agents. Configuring clusters of barebone servers. I normaly spend days on configuring things like nfs, sysctls, firewalls, upgrades, disks, crons, monitoring etc. now it's hours max. I can literally ask it to ssh into 50 vps machines and perform all tasks I tell him to do.

You mean like

  for host in vps_hosts; do
     ssh $host  errors.log 2&>1 &
  done
I imagine what you are really getting out of it is not having to type the actual commands into commands.sh?

Re: Reports of code's death are greatly exaggerated

#412

Earlier quoted context omitted.

Moors law states that the number of transistors in an integrated circuit doubles about every two years. It has nothing to say about the capabilities of statistical models. In fact in statistics we have another law which states that as you increase parameters the more you risk overfitting. And overfitting seems to already be a major problem with state of the art LLM models. When you start overfitting you are pretty mu…

Modern LLMs showed that overfitting disappears if you add more and more parameters. "Double descent" is well documented, if not well understood.

> Modern LLMs showed that overfitting disappears if you add more and more parameters.

I have not seen that. In fact this is the first time I hear this claim, and frankly it sounds ludicrous. I don‘t know how modern LLMs are dealing with overfitting but I would guess there is simply a content matching algorithm after the inference, and if there is a copyright match the program does something to alter or block the generation. That is, I suspect the overfitting prevention is algorithmic and not part of the model.

Re: Reports of code's death are greatly exaggerated

#413

> unless you believe in magic, it's only a matter of time until we reach the point at which machine intelligence is indistinguishable from human intelligence I find this flippancy about the greatest mystery in the universe extremely arrogant and incurious and wish it wouldn't be so prevalent.

Hans Moravec introduced the idea of the "landscape of human competence" , a topology representing the peaks and valleys of human capabilities. Art, writing, coding, game playing. Elevation corresponds to cognitive difficulty, and the landscape maps to everything humans are capable of doing. AI is represented as the rising waterline - when Moravec created the idea, AI was more or less constrained to a few scattered la…

Two counterpoints:

1. Implying that there are only "a few islands left" shoes a strong bias towards assuming that only thins humans do in the digital realm is relevant, when in fact, the vast majority of things humans do are not in the digital sphere at all.

2. It's pretty clear when most people say that machine intelligence is close, right now, they are alluding to LLM or Deep Learning based approaches. I don't think you should assume they mean machines will catch up in a 100 years. They seem to imply it will be by 2030 or sowmthing.

Re: Reports of code's death are greatly exaggerated

#414

Earlier quoted context omitted.

I still research efficient algorithms. You can describe these to LLMs and they do it without any prior art. They just took away the stomach churners. In fact, we probably started a communist revolution in software with anthropic/openai streaming your solutions to lesser coders.

You're actually better off using the LLM to consult textbooks from the 70s, because most likely someone already came up with a better algorithm that hasn't seen adoption yet.

Rude!

Re: Reports of code's death are greatly exaggerated

#415

Chris Lattner, inventor of the Swift programming language recently took a look at a compiler entirely written by Claude AI. Lattner found nothing innovative in the code generated by AI [1]. And this is why humans will be needed to advance the state of the art. AI tends to accept conventional wisdom. Because of this, it struggles with genuine critical thinking and cannot independently advance the state of the art. AI…

"The AI made a compiler, but it wasn't that novel, so AI is not novel" is a very poor rhetorical foundation Man - just think about what you said Two years ago that would have been beyond shocking. If 'AI is making compilers' - then that's 'beyond disruptive'. It's very true that AI has 'reversion to the mean' characteristics - kind of like everything in life .. ... but it's just unfair to imply that 'AI can't be crea…

>It's very true that AI has 'reversion to the mean' characteristics - kind of like everything in life ..

Genetic natural selection is the exact opposite mechanism of this... life is literally built around generating exception and experimenting.

Re: Reports of code's death are greatly exaggerated

#416

A week ago there was an artical about Donald Knuth asking an ai to prove something then unproved and it found the proof. I suppose it is possible that the great Knuth didn't know how to find this existing truth - but there is a reason we all doubted it (including me when I mentioned it there) i have never written a c compiler yet I would bet money if you paid me to write one (it would take a few years at least) it wo…

I don't see any reason to doubt that plausible-next-token-guessing could sometimes plausibly-next-guess a sequence that happens to decode to the answer to some question we'd not yet solved.

... it'd be even more likely if, as other have suggested in this thread, we actually had recorded the answer in writing but nobody had noticed it yet, say, but even without that I don't see why it couldn't happen.

Re: Reports of code's death are greatly exaggerated

#417

Earlier quoted context omitted.

I’ve got a “I haven’t written a line of code in one year” buddy whose startup is gaining traction and contracts. He’s rewritten the whole stack twice already after hitting performance issues and is now hiring cheap juniors to clean up the things he generates. It is all relatively well defined CRUD that he’s just slapped a bunch of JS libs on top of that works well enough to sell, but I’m curious to see the long term…

I've had to slowly and painfully learn the lesson that early on in a company's lifycycle it doesn't really matter how terrible the code is as long as it mostly works. There are of course exceptions like critical medical applications and rocket/missile guidance systems but as a general rule code quality is only a problem when it inevitably bites you much farther down the line, usually when customers start jumping ship…

> There are of course exceptions like critical medical applications and rocket/missile guidance systems but as a general rule code quality is only a problem when it inevitably bites you much farther down the line, usually when customers start jumping ship when it's obvious you can't scale or reach uptime contact targets.

My experience is it hits both new-feature velocity and stability (or the balance between those two) really early, but lots of managers don't realize that this feature that's taking literal months could have been an afternoon with better choices earlier on (because they're not in a position to recognize those kinds of things). For that matter, a lot of (greener) developers probably don't recognize when the thing that's a whole-ass project for them could have been toggling a feature flag and setting a couple config entries in the correct daemon, with better architecture, because... they don't even know what sort of existing bulletproof daemon ought to be handling this thing that somehow, horrifically, ended up in their application layer.

So the blame never gets placed where it belongs, and the true cost of half-assed initial versions is never accounted for, nor is it generally appreciated just how soon the bill comes due (it's practically instantly, in many cases).

Re: Reports of code's death are greatly exaggerated

#418
post #413

Earlier quoted context omitted.

Hans Moravec introduced the idea of the "landscape of human competence" , a topology representing the peaks and valleys of human capabilities. Art, writing, coding, game playing. Elevation corresponds to cognitive difficulty, and the landscape maps to everything humans are capable of doing. AI is represented as the rising waterline - when Moravec created the idea, AI was more or less constrained to a few scattered la…

Two counterpoints: 1. Implying that there are only "a few islands left" shoes a strong bias towards assuming that only thins humans do in the digital realm is relevant, when in fact, the vast majority of things humans do are not in the digital sphere at all. 2. It's pretty clear when most people say that machine intelligence is close, right now, they are alluding to LLM or Deep Learning based approaches. I don't thin…

To address both points - there appear to be no individual, well defined tasks that humans can do that you cannot train a machine to do. Some tasks are inefficient, uneconomical, and other impractical, but there appear to be no tasks that in principle machines cannot do. What is missing is broad generalization, human equivalent time horizons, continuous learning, and embodiment.

Robotics has passed the point of superhuman performance for any given task. Software has passed the point of superhuman performance for any given task.

Regardless of the particular technique or embodiment, the constraints aren't "is it possible in principle" but "is it too expensive" and "is this allowed by the pertinent principles and regulations and laws"

We don't have AGI that learns and adapts in real time like humans. We do have incredibly powerful algorithms that can learn from whatever data we throw at them, but many domains where it's impractical, ruinously expensive, illegal, or otherwise not possible to use AI for some other good reasons.

The few islands left to humanity are not fundamental barriers. We haven't solved intelligence, or achieved RSI or ASI or AGI yet; those were never the important thresholds.

AI has always been a question about good enough, and it looks like we've gone solidly past the good enough line into "we can probably automate everything" even if we don't solve the big problems over 5 or 10 years or beyond. I think it's very unlikely we don't solve intelligence by 2030, but even if AI stalls out where it's at right now, and all we get is the incremental improvements and engineering optimizations on current SOTA, we have enough to automate anything humans do at levels exceeding human capabilities.

What AGI and ASI do is make humans economically obsolete. Good enough AI means there might be some places where humans are needed for generalization and adaptability until the exhaustive tedious work gets done for a particular application that enables a robot or software system to be competent enough to handle the work.

Re: Reports of code's death are greatly exaggerated

#419

Earlier quoted context omitted.

In agree in principle, but the compiler is a terrible example given the amount of scaffolding afforded to the LLMs, literally hundreds of thousands of test cases covering all kinds of esoteric corners. Also (and this is coming from someone who thinks it's quite close) "AGI" is not implied by the ability to implement very-long-horizon software tasks. That's not "general" at all.

You're moving the goal posts. A year ago, _no one_ thought it could write a working compiler. Yes, the compilers we've seen today are not great. Yes, they rely too much on existing implementations. But... if you can't see which way the wind is blowing then I can't help you at this point. AGI is a meaningless milestone. No one can actually define it. The best definition I've seen is the one that ARC is using: "AI that…

What goal posts have I moved? You seem to be attributing arguments to me that I haven't made. I'm simply pointing out that the example you gave involves a level of scaffolding that most projects don't have, so that the data point is exaggerated; and that it's possible (and quite reasonable) to have an agent that is extremely good at programming while not matching what most companies and people in the space have defined as "AGI". I do believe that we'll soon have agents that can achieve Claude C Compiler–level achievements in spaces with far less scaffolding.

Re: Reports of code's death are greatly exaggerated

#420

> unless you believe in magic, it's only a matter of time until we reach the point at which machine intelligence is indistinguishable from human intelligence I find this flippancy about the greatest mystery in the universe extremely arrogant and incurious and wish it wouldn't be so prevalent.

Hans Moravec introduced the idea of the "landscape of human competence" , a topology representing the peaks and valleys of human capabilities. Art, writing, coding, game playing. Elevation corresponds to cognitive difficulty, and the landscape maps to everything humans are capable of doing. AI is represented as the rising waterline - when Moravec created the idea, AI was more or less constrained to a few scattered la…

While a 2 dimensional projection of intelligence may be a satisfying rhetorical device, I think it’s an extremely mathematically naive interpretation.

Not only is intelligence probably most accurately modeled as something extremely high dimensional, it’s probably also extremely nonlinearly traversed by learning methods, both organic and artificial. Not a topology very easily “flooded”.

In other words: bull shit.

Post reply on HN