721–730 of 1001 posts
These articles kill me. The reason LLMs (or next-gen AI architecture) is inevitably going to take over the world in one way or another is simple: recursive self-improvement. 3 years ago they could barely write a coherent poem and today they're performing at at least graduate student level across most tasks. As of today, AI is writing a significant chunk of the code around itself. Once AI crosses that threshold of con…
I think two things can be true simultaneously: 1. LLMs are a new technology and it's hard to put the genie back in the bottle with that. It's difficult to imagine a future where they don't continue to exist in some form, with all the timesaving benefits and social issues that come with them. 2. Almost three years in, companies investing in LLMs have not yet discovered a business model that justifies the massive expen…
> “most people agree that the output is trite and unpleasant to consume” That is a such a wild claim. People like the output of LLMs so much that ChatGPT is the fastest growing app ever. It and other AI apps like Perplexity are now beginning to challenge Google’s search dominance. Sure, probably not a lot of people would go out and buy a novel or collection of poetry written by ChatGPT. But that doesn’t mean the outp…
The author seems to imply that the "framing" of an argument is done so in bad faith in order to win an argument but only provides one-line quotes where there is no contextual argument. This tactic by the author is a straw-man argument - he's framing the position of tech leaders and our acceptance of it as the reason AI exists, instead of being honest, which is that they were simply right in their predictions: AI was…
Reading between the lines of the OP, the author seems to think that the future of LLMs will be determined by debate and that he can win that debate by choosing the framing of the debate. The whole meat of his article is about this debate technique, ostensibly saying that's what the other guys are doing, but really he's only described what he himself is doing.
Earlier quoted context omitted.
I'm pretty bearish on the idea that AGI is going to take off anytime soon, but I read a significant amount of theology growing up and I would not describe the popular essays from e.g., LessWrong as religious in nature. I also would not describe them as appearing poorly read. The whole "look they just have a new god!" is a common trope in religious apologetics that is usually just meant to distract from the author's o…
While it's a fair criticism, just because someone doesn't believe in a god doesn't mean the religious hardware in their brain has been turned off. It's still there and operational - I don't think it's a surprise that this hardware's attention would then be automatically tuned to a different topic. I think you can also see this in the intensification of political discussion, which has a similar intensity to religious…
AI isn't a worldview; it's an extremely powerful tool which some people happen to be stronger at using than others, like computers or fighter jets. For people who empirically observe that they've been successful at extracting massive amounts of value from the tool, it's easy to predict a future in which aggregate economic output in their field by those who are similarly successful will dwarf that of those who aren't. For others, it's understandable that their mismatched experience would lead to skepticism of the former group, if not outright comfort in the idea that such productivity claims are dishonest or delusional. And then of course there are certainly those who are actually lying or deluded about fitting in the former group.
Every major technology or other popular thing has some subset of its fandom which goes too far in promotion of the thing to a degree that borders on evangelical (operating systems, text editors, video game consoles, TV shows, diets, companies, etc.), but that really has nothing to do with the thing itself.
Speaking for myself, anecdotally, I've recently been able to deliver a product end-to-end on a timeline and level of quality/completeness/maturity that would have been totally impossible just a few years ago. The fact that something has been brought into existence in substantially less time and at orders of magnitude lower cost than would have been required a few years ago is an undeniable observation of the reality in front of me, not theological dogma.
It is, however, a much more cognitively intense way to build a product — with AI performing all the menial labor parts of development, you're boxed into focusing on the complex parts in a far more concentrated time period than would otherwise be required. In other words, you no longer get the "break" of manually coding out all the things you've decided need to be done and making every single granular decision involved. You're working at a higher level of abstraction and your written output for prompting is far more information-dense than code. The skills required are also a superset of those required for manual development; you could be the strongest pre-LLM programmer in the world, but if you're lacking in areas like human language/communication, project/product management, the ability to build an intuition for "AI psychology", or thinking outside the box in how you use your tools, adapting to AI is going to be a struggle.
It's like an industry full of mechanics building artisan vehicles by hand suddenly finding themselves foisted with budgets to design and implement assembly lines; they still need to know how to build cars, but the nature of the job has now fundamentally changed, so it's unsurprising that many or even most who'd signed up for the original job would fail to excel in the new job and rationalize that by deciding the old ways are the best. It's not fair, and it's not anyone's fault, but it's important for us all to be honest and clear-eyed about what's really happening here. Society as a whole will ultimately enjoy some degree of greater abundance of resources, but in the process a lot of people are going to lose income and find hard-won skills devalued. The next generation's version of coal miners being told to "learn to code" will be coders being told to "learn to pilot AI".
I think two things can be true simultaneously: 1. LLMs are a new technology and it's hard to put the genie back in the bottle with that. It's difficult to imagine a future where they don't continue to exist in some form, with all the timesaving benefits and social issues that come with them. 2. Almost three years in, companies investing in LLMs have not yet discovered a business model that justifies the massive expen…
But have we ever had a general purpose technology (steam engine, electricity) that failed to change society?
Earlier quoted context omitted.
A few things that are here: - LLMs - Cryptocurrencies - Mobile phones Neither are going away, all are part of our future, but not equally. The inevitabilism argument is that cryptocurrencies were just as hyped a few years ago as LLMs are now, and they're much less here now. So if you have an objection to LLMs being hyped and not wanting to use them, there's a real case they may slide into the background as a curious…
Crypto was absolutely not where LLM is now. It's historical revisionism.
Earlier quoted context omitted.
The IT industry is also full of salesmen and con men, both enjoy unrealistic exaggeration. Your statements would not be out of place 20 years ago when the iPhone dropped. Your statements would not be out of place 3 years ago before every NFT went to 0. LLMs could hit an unsolvably hard wall next year and settle into a niche of utility. AI could solve a lengthy list of outstanding architectural and technical problems…
> LLMs could hit an unsolvably hard wall next year and settle into a niche of utility LLMs in their current state have integrated into the workflows for many, many IT roles. They'll never be niche, unless governing bodies come together to kill them. > I can't find a single open source codebase, actively used in production, and primarily maintained and developed with AI Straw man argument - this is in no way a metric…
Having a niche is different from being niche. I also strongly believe you overstate how integrated they are.
> Straw man argument - this is in no way a metric for validating the power of LLMs as a tool for IT roles. Can you not find open source code bases that leverage LLMS because you haven't looked, or because you can't tell the difference between human and LLM code?
As mentioned, I have looked. I told you what I found when I looked. And I've invited others to look. I also invited you. This is not a straw man argument, it's making a prediction to test a hypothesis and collecting evidence. I know I am not all seeing, which is why I welcome you to direct my eyes. With how strong your claims and convictions are, it should be easy.
Again: You claim that AI is such a productivity boost that it will rock the IT industry to its foundations. We cannot cast our gaze on closed source code, but there are many open source devs who are AI-friendly. If AI truly is a productivity boost, some of them should be maintaining widely-used production code in order to take advantage of that.
If you're too busy to do anything but discuss, I would instead invite you to point out where my reasoning goes so horrendously off track that such examples are apparently so difficult to locate, not just for me, but for others. If one existed, I would additionally expect that it would be held up as an example and become widely known for it with as often as this question gets asked. But the world's full of unexpected complexities, if there's something that's holding AI back from seeing adoption reflected in the way I predict, that's also interesting and worth discussion.
You're my blog hero, thank you for being cool and setting a good example. Also really important LLM hype reminder.
One of the negative consequences of the “modern secular age” is that many very intelligent, thoughtful people feel justified in brushing away millennia of philosophical and religious thought because they deem it outdated or no longer relevant. (The book A Secular Age is a great read on this, btw, I think I’ve recommended it here on HN at least half a dozen times.) And so a result of this is that they fail to notice t…
I'm pretty bearish on the idea that AGI is going to take off anytime soon, but I read a significant amount of theology growing up and I would not describe the popular essays from e.g., LessWrong as religious in nature. I also would not describe them as appearing poorly read. The whole "look they just have a new god!" is a common trope in religious apologetics that is usually just meant to distract from the author's o…
I mean the lack of self awareness you have here is amazing.
One of the negative consequences of the “modern secular age” is that many very intelligent, thoughtful people feel justified in brushing away millennia of philosophical and religious thought because they deem it outdated or no longer relevant. (The book A Secular Age is a great read on this, btw, I think I’ve recommended it here on HN at least half a dozen times.) And so a result of this is that they fail to notice t…
Techno Calvinists vs Luddite Reformists is a very funny image. Agree - although it's an interesting view, I think it's far more related to a lack of idealogy and writing where this has emerged from. I find it more akin to a distorted renaissance. There's such a large population of really intelligent tech people that have zero real care for philisophical or religious thought, but still want to create and make new thin…
He tells a history of them constantly failing because the core ideology of “cybernetics” is underlying them all and fails to be adaptive enough to match our DNA/Body/mind combined cognitive system. Especially when scaled to large groups.
He makes the second point that humanity and many thinkers constantly also resort to the false notion of “naturalism” as the ideal state of humanity, when in reality there is no natural state of anything, except maybe complexity and chaos.