pron

21 ms·

271.
by pron 6mo ago
I agree with both statements, but that doesn't change the problem I stated. If an agent produces reasonable code 80-90% of the time, and 10-20% of the time it makes mistakes that could render the codebase irretrievably unevolvable once
272.
by pron 6mo ago
> The invariant, stated informally, would be hard to prove is broken by a human reviewer in the loop. Spoken language isn’t precise enough for the task. That depends on the invariant. Some are behavioural, like "variable x must be e
273.
by pron 6mo ago
Sure, if you carefully review the agent's output, including tests, you can get good results. If you don't carefully review the output, you obviously have no idea if it's good enough for you. The only way to find out is that 3
274.
by pron 6mo ago
That doesn't quite work, and precisely for the reason I mentioned: You can definitely tell the AI to follow some strategy, but at some point the strategy will need to change, and the AI won't tell you that (even if you tell it to)
275.
by pron 6mo ago
Sure. I'm talking about production software that needs to survive and evolve for a long while.
276.
by pron 6mo ago
It's not enough to enforce the invariants because they may need to change. You need to follow the invariants when they're right, and go back and reconsider them when they prove unhelpful. Knowing which is the case requires judgmen
277.
by pron 6mo ago
Not quite, because the architecture often needs to evolve when you learn more as the project evolves. People will complain when they feel the constraints drive them to unnatural workarounds, the agents don't. You can try telling the ag
278.
by pron 6mo ago
Yep. The only people I've heard saying that generated code is fine are those who don't read it. The problem is that the mitigations offered in the article also don't work for long. When designing a system or a component we ha
279.
by pron 6mo ago
The point of TLA+ is easily readable specs with refinements. That's why it was designed in a novel way rather than in the older executable-with-more-complex-logic style that Lean (and other maths/spec languages) offer. I'm no
280.
by pron 6mo ago
> There doesn't seem to be support for your claim that: "Having more domain coverage is easier and more valuable…". Since programming languages specifically optimise differently for the different conditions I listed above,
281.
by pron 6mo ago
How do you have "aggressive error detection" when one of the most common and pernicious mistakes agents make are architectural? The behaviour is fine, but the code is overly defensive, hiding possible bugs and invariant violations
282.
by pron 6mo ago
> The chickens will always come home to roost: eventually the memory limit will be reached and you'll have to pay that collection cost in full. This sounds like you don't understand how GCs work. I can't cover all the basi
283.
by pron 6mo ago
> So where are the examples of that being done? Whenever people want to get valuable information. As I said, we in OpenJDK have a couple hundred benchmarks, some macro, many micro, which are meant to give a decent coverage of the things
284.
by pron 6mo ago
Oh, I think you just haven't read my comments. What I wrote, and I quote was: "AI can help software developers in many ways, but not like that." I was saying how much more productive LLMs make developers unless you use them
285.
by pron 6mo ago
Nah, it's a skill issue on the part of those who believe in "agent swarms" (in fact, that's how I recognise AI noobs; they think swarms work). Studies (like this [1]) and Anthropic's experiements have told us they d
286.
by pron 6mo ago
They're not productive for any workflow is my point because they don't produce sustainable software, yet that's exactly what Armstrong is calling for. They don't work, and people experienced with AI workflows already k
287.
by pron 6mo ago
If you carefully review the code then you're not doing what Armstrong was talking about. If you're not reviewing the code, then you don't really know what it is that the AI built. Of course it passes tests; that's not th
288.
by pron 6mo ago
I was being sarcastic.
289.
by pron 6mo ago
I use GPT 5.5. Sometimes it does what you say. It certainly finds silly mistakes in my code better than I could. But frequently enough it makes catastrophic architectural mistakes in its own code.
290.
by pron 6mo ago
The only way you could possibly know that is if you're reviewing the code, which means you're not "managing fleets of agents". If you're not reviewing the code (and you wouldn't be if you're managing fleet
291.
by pron 6mo ago
I don't know how anyone who carefully and closely reviews their output could possibly think that. Much of the time their code is fine, but every now and again they make a catastrophic (though often well-hidden) mistake that is so bad t
292.
by pron 6mo ago
If you carefully read the agent's output you'll see why. It adds layers upon layers of workarounds and defences that hide serious problems, until the codebase reaches a point where the agent can no longer understand it and work wi
293.
by pron 6mo ago
If the code is really simple, it's cheap to read it. When people don't read it (and when they need to use "fleets of agents"), it's because it's not so simple, and then the people who trust the outcome are thos
294.
by pron 6mo ago
It's not about caring how it works. It's about caring that it keeps working at all even after you add stuff to it for a year or three (and nearly all software written by companies is software they evolve).
295.
by pron 6mo ago
Except people who are learning to leverage it appropriately already know better than to generate important production code by "managing fleets of agents".
296.
by pron 6mo ago
And if the trend doesn't continue? I understand that a company with Coinbase's performance has little to lose and not many options, but many companies are in a better position. The problem is that executives could take the 15-20%
297.
by pron 6mo ago
> If you’re more or less experienced, you can easily see the “good” and “bad” sides of it. So you kinda plan it out in a way that you can “evolve AI generated software”. If you're truly "managing fleets of agents" there&#x
298.
by pron 6mo ago
> Over the past year, I’ve watched engineers use AI to ship in days what used to take a team weeks. No, you didn't. You watched engineers use AI to ship in days something that looks like what used to take a team weeks. After enoug
299.
by pron 6mo ago
I think you're referring to the less important point I made. Correcting for apples-to-apples is harder and less valuable. Having more domain coverage is easier and more valuable (especially since the current coverage is so narrow and
300.
by pron 6mo ago
Ok, but if the measurements were wider in scope they could at least offer a more interesting, well-rounded, and perhaps even relevant basis for discussion (even if the other flaws, which are harder to fix, remained).

More ›