Thats because humans have "understanding" they can use to assess quality, without understanding "trying harder" just means spending more "effort" distilling an average result, at best over a larger sample size.
yeah, I was wondering the same.. kinda big deal that they first of all had the issue and second, that they fixed it in January and only now message about it
His anti-LLM stance isnt unsurprising given the values he's posted about regularly for years, if anything his ability to avoid cargo-cult/hype nonsense is why I (and seemingly others) value his writing so much.
Isn't a huge part of the point of developing in the open so that people can assess the entire technology, including how its developed, and choose if they want to use it?
Folks using LLMs will have to accept that that will turn people away from using their software ¯\_(ツ)_/¯
I always find folks bringing up rubber ducking as a thing LLMs are good at to be misguided. IMO, what defines rubber ducking as a concept is that it is just the developer explaining what their doing to themselves. Not to another person, and not to a thing pretending to be a person. If you have a "two way" or "conversational" debugging/designing experience it isnt rubber ducking, its just normal design/debugging.
The moment I bring in a conversational element, I want a being that actually has problem comprehension and creativity which an LLM by definition does not.
Sometimes I don't want creativity though, I'm just not familiar enough with the solution space and I use the LLM as a sort of gradient descent simulator to the right solution to my problem (the LLM which itself used gradient descent when trained, meta, I know). I am not looking for wholly new solutions, just one that fits the problem the best, just as one could Google that information but LLMs save even that searching time.
I've read the thread and in my mind you're missing that LLMs increase the surface area of visibility of a thing. It's a probe. It adds known unknowns to your train of thought. It doesn't need to be "creative" about it. It doesn't need to be complete or even "right". You can validate the unknown unknown since it is now known. It doesn't need to have a measured opinion (even though it acts as it does), it's really just topography expansion. We're getting in the weeds of creativity and idea synthesis, but if something is net-new to you right now in your topography map, what's so bad about attributing relative synthesis to the AI?
Honest, non-confrontational, non-passive aggressive question: Have you used any of the latest models in the last 6 months to do coding? Or frankly, in the last year?
What are you defining as free versus frontier, and for what purpose? For coding there is a big difference between Opus and GPT 5.3/4 versus Sonnet and other models such as open weight ones.
I just don't use the web much anymore because the experience has degraded so much over the past several years and it has become decreasingly useful at work as well. I do sometimes need to search for a document and find Kagi pretty good for that, but the old way of using a search engine to kind of explore and discover stuff just isn't viable anymore, unfortunately.
I administer software for a living so I read a lot of documentation of that software but it comes with the software so I don't ever really need to search for it; I also read and participate in some forums and us the relevant IRC channels.
If you're not familiar with the problem space, by definition you don't know whether or not that's the case. The problem spaces I do know well, I know the LLM isn't good at it, so why would I assume it's better at spaces I don't know?
I said familiar enough, not familiar. For example, let's say I'm building an app I know needs caching, the LLM is very good at telling me what types of caching to use, what libraries to use for each type, and so on, for which I can do more research if I really want to know specifically what the best library out of all the rest are, but oftentimes its top suggestion is, like I said, good enough for my purpose of e.g. caching.
I still don't get what you're saying. If you possess enough information to accurately judge the LLM's suggestions you possess enough information to decide on your own. There's not really a way around that.
Of course I'm deciding on my own, I'm not letting the LLM decide for me (although some people do). But the point is whatever the suggestion is is merely an implementation detail that either solves my problem or not, not sure what part of that is confusing. Replace LLM with glorified Google and maybe it's less confusing.
Theoretically the LLM would weight more popular suggestions more too. Regardless you're reading too much into this, either use the LLM or don't, I'm not sure if someone else can convince you. As I said for my purposes of getting shit done it works perfectly fine and works more like a research tool than anything else, especially if it can understand my specific use case unlike general research tools like Google or Stack Overflow.
If you don't review the code it generates then that's still on you. There isn't an excuse for handing in breaking PRs like your juniors. It's a tool at the end of the day and it's the responsibility of the user to utilize it correctly.
This is a very strange and contradictory situation. I'm not sure there's any point in engaging with you since there is nothing but a stream of weak dismissals farming for engagement.
You dismiss LLMs because of factual inaccuracy, which is fair, but now you're doubling down on an anti search engine stance, which is weird, because the modern substitute is letting LLMs either use search engines on your behalf or learn the entire internet with some error and you've dismissed both.
Yes, I'm the "backwards" guy who still uses search engines. We still exist.
I've noticed that HN can attract some of the most extreme people I've ever seen, and I suppose there is precedent in the tech world when I'm reminded of the story of Stallman not using a browser but instead sending webpages to his email where he then reads the content. It's literally nonsensical for 99.9999% of the population and I've read similar absurd things on HN as well.
This person not using LLMs is fine, I understand the argument like you said, but the double down on not using search engines either makes me not take anything they say seriously. Not to be too crass but it reminds me of this situation on the nature of arguing on the internet [0].
Absolutely, the whole point of the rubber duck is that it's inanimate. The act of talking to the rubber duck makes you first of all describe your problem in words, and secondly hear (or read) it back and reprocess it in a slightly different way. It's a completely free way to use more parts of your brain when you need to.
LLMs are a non-free way for you to make use of less of your brain. It seems to me that these are not the same thing.
Maybe it’s just a semantic distinction, which, sure. I guess I’d just call it research? It’s basically the “I’m reading blogs, repos, issue trackers, api docs etc. to get a feel for the problem space” step of meaningful engineering.
But I definitely reach for a clear and concise way to describe that my brain and fingers are a firewall between the LLM and my code/workspace. I’m using it to help frame my thinking but I’m the one making the decisions. And I’m intentionally keeping context in my
brain, not the LLM, by not exposing my workspace to it.
Sometimes people just need something else to tell them their ideas are valid. Validation is a core principle of therapeutic care. Procrastination is tightly linked to fear of a negative outcome. LLMs can help with both of these. They can validate ideas in the now which can help overcome some of that anxiety.
Unfortunately they can also validate some really bad ideas.
I feel I've had the most success with treating it like another developer. One that has specific strengths (reference/checklists/scanning) and weaknesses (big picture/creativity). But definitely bouncing actual questions that I would say to a person off it.
The best counter to this is to just forego these tools and develop actual knowledge. Knowledge & skill is still power, no matter what the AI propagandists say
Removing AI quite clearly doesn't fit into that category. It doesnt reduce human suffering, hell the primary propagandists behind it are explicitly anti-humanists and want it to replace human creativity. Thankfully its awful at that
reply