20k

4 ms·

31.
by 20k 1mo ago
Shockingly poor security to let an application have totally unrestricted access to the web with no review, of course this kind of stuff is going to happen
32.
by 20k 1mo ago
I mean, they've already violated the law in acquiring all their training data already, why would they be uncomfortable violating a contract to get more training data?
33.
by 20k 1mo ago
I mean the models have literally trained on: 1. Child porn 2. Stolen music 3. Private github repos, before that was 'stopped' 4. Illegally pirated books Them training on company prompts against the terms of service would be one of
34.
by 20k 1mo ago
It isn't, in both cases you're protected by exactly the same legal system
35.
by 20k 1mo ago
You have to be rather naive if you don't think these companies don't simply train on your prompts with or without your consent. They literally scrape everything - legal or not - and claim its fair use to train on, including straig
36.
by 20k 1mo ago
These companies absolutely do have people who's job it is to sit on social media and promote these products, its marketing 101
37.
by 20k 2mo ago
Yes, it strongly suggests that AI as a learning tool is at best pretty much useless, but on average its strongly detrimental to most people exposed to it as it makes it too easy to cheat
38.
by 20k 2mo ago
They'll all be getting around to this sooner or later, its just too expensive
39.
by 20k 2mo ago
Extremely silly. Even if LLMs did everything that everyone says they do (which they absolutely don't), they still hit a fundamental limit of complexity when they stop being useful Its only a good idea if you work in selling tokens, oth
40.
by 20k 2mo ago
This doesn't seem to match the claims of 10x productivity that are being levied though. In fact it seems like the overall productivity is lower
41.
by 20k 2mo ago
Then why isn't it showing up in open source at all? The people who file PRs are just regular average devs, but the quality of LLM PRs seems to be universally crap in comparison
42.
by 20k 2mo ago
>I think it would be reasonable to ask which major open source projects aren't using AI assistance? Where are all the vast quantity of LLM commits for these open source projects, if these engineers are truly 10xing their productivit
43.
by 20k 2mo ago
How is it challenging for open source projects? You just write a PR, and try to get it to meet the maintainer's quality bar. That's exactly how all code is written in any environment - you do code review and refine a PR until it
44.
by 20k 2mo ago
If they were sneaking LLM generated code into these projects, we'd expect to see their feature development 10x even if they covertly aren't telling anyone. Which open source codebases have seen a truly massive 10x increase in deli
45.
by 20k 2mo ago
You'd think one 10x engineer or one core developer already part of the project would have completely eclipsed all the other engineers, and worked around the difficulties there Or, failing that, you'd think any of these engineers
46.
by 20k 2mo ago
They don't though, very few people were anti AI in the early days. Open source projects were largely initially very interested in LLM usage, and generally had a pro AI contribution policy. This changed when it did not help their lives
47.
by 20k 2mo ago
There's no reason that some people wouldn't be using their new found 10x superpowers on open source projects though
48.
by 20k 2mo ago
The the idea that engineers are covertly sneaking huge quantities of LLM generated code into these projects doesn't pass the smell test, as there's lots of evidence that this isn't true. It'd be incredibly obvious if new
49.
by 20k 2mo ago
Its always been possible to trade long term productivity for short term gains with technical debt. This is why the bar I'm interested in is long term projects, which have proven to have long term success, instead of a small disposable
50.
by 20k 2mo ago
Sure, it just seems a little odd that no LLMgineer ever contributes their incredible more performant scheduler back though right? After all if you can do it with claude, anyone can, all it'd take is to ask claude to rewrite it. Linux a
51.
by 20k 2mo ago
If that were true, we'd expect to see massively accelerated velocity of open source projects by these engineers. They should be creating new open source projects at a truly astounding rate, with new tooling springing up every day that
52.
by 20k 2mo ago
Why would the PR format be bad for LLMs? Its just code review + merging in branches, which is what every company should be doing anyway Nothing's stopping core developers from adopting LLM generated code for themselves, while banning i
53.
by 20k 2mo ago
I recently asked it a series of technical physics questions about a field I used to be involved in to evaluate whether it'd be useful for doing research, and literally every single answer it gave was critically wrong - even for basic q
54.
by 20k 2mo ago
>If agentic development actually worked the way any of them say it does I think its fascinating just how much of a gap there is between what's being claimed, and the verifiable observable data of the open source world. Major open so
55.
by 20k 2mo ago
I don't think there's a way to make it secure while still permitting it unprompted external access Eg: Any web request is a security vulnerability, there's no way to do it if the web requests are being made maliciously Say th
56.
by 20k 2mo ago
>So why are you pretending this is something new or novel? Because as the OP of this thread and LLM users are rediscovering, that understanding is the most critical part of software development >I don't think this personal opinio
57.
by 20k 2mo ago
Its weird how absolutely desperate we are to avoid humans actually using their brains at any step of the coding process. I don't know if we just need to tell people that its fine to need to think about things, and that its actively hea
58.
by 20k 2mo ago
heaven forbid we use our meat to actually completely understand it in the first place
59.
by 20k 2mo ago
LLM speedups have always been predicated on not understanding the code they produce as well as if you wrote it yourself. Reviewing even well written code to understand it in depth, has always taken longer than simply writing it yourself - l
60.
by 20k 2mo ago
People have tried all kinds of techniques for VLIW, including techniques that are much better than a purpose built AI, AI isn't a magic silver bullet. Fundamentally there's no reason you can't analyse a piece of code to dea

More ›