991–1000 of 1001 posts
[flagged]
I think it is unethical of you to post comments like this one without disclosing you have no expertise to judge and personal stakes that make you desire the article to be wrong.
So many people putting expectations up to knock down about models. Infinite reasons to critique them. Please dispense with anyone's "expectations" when critiquing things! (Expectations are not a fault or property of the object of the expectations.) Today's models (1) do things that are unprecedented. Their generality of knowledge, and ability to weave completely disparate subjects together sensibly, in real time (and…
Are they progressing quickly? Or was there a step-function leap about 2 years ago, and incremental improvements since then? I tried using AI coding assistants. My longest stint was 4 months with Copilot. It sucked. At its best, it does the same job as IntelliSense but slower. Other times it insisted on trying to autofill 25 lines of nonsense I didn't ask for. All the time I saved using Copilot was lost debugging the…
Thia might appear to ba a shallow answer but I do not think it is. AI has taken a very long road from early conceptions, by Turing and others, to a tool whose value we can argue about, but which is getting attention and use everywhere.
The mere fact that "are they progressing rapidly" is a question, is a testament to an incredible uptick in speed of progression.
"Is AI progressing quickly?" is the new "Are we there yet?"
Earlier quoted context omitted.
It makes engineers a hell of a lot more efficient and opens up software to a whole new class of people. There is plenty of data to back this up. Based on that alone it’s worth quite a lot.
I don't doubt the first part, but how true is the second? Is there a shortage of React apps out there that companies are desperate for? I'm not having a go at you--this is a genuine inquiry. How many average people are feeling like they're missing some software that they're able to prompt into existence? I think if anything, the last few years have revealed the opposite, that there's a large/huge surplus of people in…
I have a friend in IT who never knew how to code but is now saving a lot of time every month with just basic scripts.
Earlier quoted context omitted.
The problem Sabine tries to communicate is that reality is different from what the cash-heads behind main commercial models are trying to portray. They push the narrative that they’ve created something akin to human cognition, when in reality, they’ve just optimised prediction algorithms on an unprecedented scale. They are trying to say that they created Intelligence, which is the ability to acquire and apply knowled…
You lay the rhetoric on so thick (“cash heads”, “pushing the narrative”, “corporate overlords”, “profit-making abomination”) that it’s hard to understand your actual claim. Are you trying to say that LLMs are useful now but you think that will stop being the case at some point in the future?
The tech industry, especially big corporations, doesn’t chase innovation; it chases repeatable, predictable profit.
Earlier quoted context omitted.
The problem Sabine tries to communicate is that reality is different from what the cash-heads behind main commercial models are trying to portray. They push the narrative that they’ve created something akin to human cognition, when in reality, they’ve just optimised prediction algorithms on an unprecedented scale. They are trying to say that they created Intelligence, which is the ability to acquire and apply knowled…
If it's just cash-heads pushing a narrative, where do Bengio and Hinton fit? Stuart Russell? Doug Hofstadter? I mean fine, argue that they're mistaken to be concerned, if that's your belief. But dismissing it all as obvious shilling is not that argument.
I'm not a functionalist and my belief is that AI — especially LLMs — will never achieve real understanding or consciousness, no matter how much we scale them. Language prediction is just a computation, but human thought is more than that.
Earlier quoted context omitted.
Are they progressing quickly? Or was there a step-function leap about 2 years ago, and incremental improvements since then? I tried using AI coding assistants. My longest stint was 4 months with Copilot. It sucked. At its best, it does the same job as IntelliSense but slower. Other times it insisted on trying to autofill 25 lines of nonsense I didn't ask for. All the time I saved using Copilot was lost debugging the…
have you tried it recently? o3-mini-high is really impressive. If you ease into talking to it about your intent and outlining the possible edge and corner cases it will write nuanced rust code 1000 lines at a time no problem
Earlier quoted context omitted.
If it's just cash-heads pushing a narrative, where do Bengio and Hinton fit? Stuart Russell? Doug Hofstadter? I mean fine, argue that they're mistaken to be concerned, if that's your belief. But dismissing it all as obvious shilling is not that argument.
If we're calling out names, what about Roger Penrose, John Searle, Stuart Hameroff, Hubert Dreyfus, Henry Stapp? These are very intelligent people and I suggest to get acquainted with their work. Neural scaling laws are real and no matter if you put into training 10e-8 petaFLOP/days of compute or 200000 petaFLOP/days you will hit irreducible error constant at Efficient Compute Frontier. I'm not a functionalist and my…
Above you wrote "we all know the only real Intelligence ... is" as your support for attributing venial motives to people taking AI progress seriously. OK, now I know your basis for that claim. I've read three of the guys you mention, agree they're intelligent and except for Searle have some good things to say. But it's really unconvincing as support for an AI-is-fake claim, and especially for an everyone-knows claim.
Earlier quoted context omitted.
It really depends on the task. Like Sabine, I’m operating on the very frontier of a scientific domain that is extremely niche. Every single LLM out there is worse than useless in this domain. It spits out incomprehensible garbage. But ask it to solve some leet code and it’s brilliant.
The question I ask afterwords then is: is solving some leet code brilliant? Is designing a simple inventory system brilliant if they've all been accomplished already? My answer tends towards no, since they still make mistakes in the process, and it harms newer developers from learning.
Earlier quoted context omitted.
> What's worse, people are treating them as authoritative. … I've both seen online and heard people quote LLM output as if it were authoritative. Thats not an LLM problem. But indeed quite bothersome. Dont tell me what Chatgpt told you. Tell me what you know. Maybe you got it from ChatGPT and verified it. Great. But my jaw kind of drops when people cite an LLM and just assume it’s correct.
> But my jaw kind of drops when people cite an LLM and just assume it’s correct. Yes but they're literally told by allegedly authoritative sources that it's going to change everything and eliminate intellectual labor, so is it totally their fault? They've heard about the uncountable sums of money spent on creating such software, why would they assume it was anything short of advertised?
Well... yeah.
Earlier quoted context omitted.
blockchain is here to stay, dont kid yourself
While it could be said that LLMs are in the 'peak of inflated expectations', blockchain is definitely still in the 'trough of disillusionment'. Even if there was a way for blockchain to affordably facilitate everyday transactions without destroying the planet and somehow sideloading into government acceptance, it's not clear that there would be anything novel enough to motivate people to use it vs a bank - beyond a c…