Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Fable and the end of the free lunch

dbreunig.com

Post reply on HN
Print view View original

261–268 of 268 posts

Re: Fable and the end of the free lunch

#261

Post by 59nadir »

ShinyLeftPad wrote: post #256
59nadir wrote: post #249

Earlier quoted context omitted.

I haven't said anything about "good", I just pointed out that there isn't exactly an alternative to Chinese models if censorship and restrictions are regarded as bad for longevity. You can't really do better than the Chinese models for longevity; US models are by far the worst in this regard. So "What about censorship?" is an absolutely hilarious question to ask when Chinese models are presented as an alternative. Ye…

Censorship or guardrails? "nothing happens in 1989" is censorship. "I won't tell you how to build a bioweapon for genocide" is guardrails. I like the second one because I like to be alive.

I don't share your enthusiasm for the guardrails and I don't believe at all that they accomplish what they're supposedly created for.

Re: Fable and the end of the free lunch

#262

Post by misalliance »

enraged_camel wrote: post #5

>> GLM 5.2 is worth focusing on. It came out the same week as Fable and is roughly 1/9th the cost (and ~1/5th the cost of Opus 5). Is GLM 1/9th the quality of Fable? Perhaps, for certain classes of tasks. But for most rote coding it’s more than sufficient. Especially when provided with great context. I frequently chat with Fable to interrogate and shape a design, before handing off a brief to GLM. People say stuff li…

[dead]

Re: Fable and the end of the free lunch

#263

Post by saghm »

lern_too_spel wrote: post #259
saghm wrote: post #253

Earlier quoted context omitted.

And my point is that I've never met a human who confidently asserts incorrect information in such a broad range of domains rather than just admitting that they don't know

The point is that it's trivial to fix the problem of cooking the waterfowl or writing your code with existing tools. Just ask them to verify.

If you have to say "don't make up something" for every possible question you ask in order for it not to make up something, that's a massive usability issue for regular people. If saying "don't make up something" will still result in it making up something up some of the time, that also might be a massive usability issue depending on if "some of the time" means 0.0001% or 1%.

Having a natural language interface where you need to go out of your way to specify that you want an accurate answer rather than just a plausible one defeats the entire purpose of it being a natural language interface for normal people. In certain professional contexts, it can be useful, but I don't buy it at all that it makes sense to ask everyone in their everyday lives to go out of their way to specify that they actually want correct answers to their questions.

Re: Fable and the end of the free lunch

#264

Post by matheusmoreira »

ShinyLeftPad wrote: post #242
rrr_oh_man wrote: post #150

Earlier quoted context omitted.

Everyone censors for their core jurisdiction/audience. The Enlightened West just calls this guardrails

"nothing happens in 1989" is censorship. "no I won't tell you how to build a bioweapon for genocide" is guardrails. idk about you but i WANT the second thing, because I like to be alive.

> "no I won't tell you how to build a bioweapon for genocide" is guardrails.

I call that censorship too. I'm curious enough that I want to know about such things. I don't want any limitations on what I'm allowed to understand and know about.

Re: Fable and the end of the free lunch

#265

Post by lern_too_spel »

saghm wrote: post #263
lern_too_spel wrote: post #259

Earlier quoted context omitted.

The point is that it's trivial to fix the problem of cooking the waterfowl or writing your code with existing tools. Just ask them to verify.

If you have to say "don't make up something" for every possible question you ask in order for it not to make up something, that's a massive usability issue for regular people. If saying "don't make up something" will still result in it making up something up some of the time, that also might be a massive usability issue depending on if "some of the time" means 0.0001% or 1%. Having a natural language interface where…

> If you have to say "don't make up something" for every possible question you ask

You don't. It goes in the system prompt.

Re: Fable and the end of the free lunch

#266

Post by eru »

somenameforme wrote: post #245
josephg wrote: post #142

Earlier quoted context omitted.

Maybe - we'll have to wait and see. By my reckoning, there's a significant chance most software engineers will be unemployable within a few years. But I'm not 100% confident that there'll be a utopia waiting for us, as an alternative.

Ostensibly there is no reason this isn't already happening at large companies. I used to think it was an expertise bias, since the impact of LLMs is inversely proportional to your own ability - but I think this is also no longer the case. Even if you're extremely skilled, they can already greatly supplement, if not supplant, you on implementation and increasingly even planning/architecture. So my new pet theory is th…

> So my new pet theory is that this is one of the few times we're seeing a positive effect from the heads of all of these big businesses being part of weird public (e.g. WEF) and private (e.g. Bohemian Club) orgs where they get to together and conspire to conquer the world or whatever.

I doubt that conspiracy theory.

Re: Fable and the end of the free lunch

#267

Post by somenameforme »

eru wrote: post #266
somenameforme wrote: post #245

Earlier quoted context omitted.

Ostensibly there is no reason this isn't already happening at large companies. I used to think it was an expertise bias, since the impact of LLMs is inversely proportional to your own ability - but I think this is also no longer the case. Even if you're extremely skilled, they can already greatly supplement, if not supplant, you on implementation and increasingly even planning/architecture. So my new pet theory is th…

> So my new pet theory is that this is one of the few times we're seeing a positive effect from the heads of all of these big businesses being part of weird public (e.g. WEF) and private (e.g. Bohemian Club) orgs where they get to together and conspire to conquer the world or whatever. I doubt that conspiracy theory.

On what account? It's certainly true that these people are part of these orgs. And it's also certainly true that they're discussing 'mental automation' (as a catch-all for LLM stuff) and its economic/social consequences amongst themselves. And finally it's also true that they're not really acting as much on LLMs as they ostensibly could, when normally they'd do pretty much anything, regardless of longer term consequences, if it'd increase next quarter's margins by a point or two.

I think the only assumption that's meaningfully debatable is whether the current SOTA are able to supplant labor to a more significant degree or not. And while I suppose that's going to inextricably remain an opinion, I think it's reasonably objective to say that hallucination rates have sharply declined, and overall code quality/coherence is sharply up.

Re: Fable and the end of the free lunch

#268

Post by Tuna-Fish »

poincareball wrote: post #250
Tuna-Fish wrote: post #75

Earlier quoted context omitted.

Please explain why you think cheaper/faster is not coming? All current devices used to run AI are very far from an efficient solution to the problem. What you really want is a pure dataflow architecture, instead of a von Neumann machine. The reason people aren't really making them yet is that when you build one, even if you use SRAM for the weights, you are binding yourself to the dimensions of the model you target -…

And how are we going to build these things with current limitations on fabrication? There are only a few places in the world building 3nm chips. And they're under serious threat of foreign violence. It's highly likely that over the next 10 years we find demand and loss of production further constrains supply. Designing a specific model into silicon sounds like one of the worst possible ideas. No better way to freeze…

Designing a specific model into silicon buys you three orders of magnitude of speed and energy-efficiency. Flexible, software-defined solutions are used today because adaptability is key in an environment where people expect better models in the very near future that would obsolete the expensive investment in masks. If this situation ends, and people stop expecting better models, models will be directly etched into silicon and adaptable systems will not be able to compete with them.
Post reply on HN