@ylecun@pabloestevezh@huxlab 😅👏Thank you Yann, i feel like whenever i listen to your arguments, even here on X, or talks, that really youre the only one with a common sense, that actually CAN see the forest for the trees.
And it should´nt be that way i guess... Please more common sense @ academicworld
Dario's spiel since GPT-2 has always been about hyping the dangers so that open models would be made illegal and *he* would stay in control.
He left OpenAI because he wasn't going to have control.
It's like the Ottoman Empire banning the printing press for Arabic so they would keep in control of the dogma.
It stops the population from freely accessing diverse knowledge.
It's a new form of medieval obscurantism.
@anshuc No they are super goated for almost anything, expect, UX and UI, "beauty" and tricking you into believing, the model does "infer" more than it does, by serving you a bigger variance of high quality slop, learned in the RL pipeline, like Fable does.
@KactuzY@MasterBayesian@thsottiaux Yeah i see Fable and Antrophic models in general as high quality, fancy slop.
I mean, GPT slops too, but 6.2 not as much anymore. It wants you to actually deliver complex tasks and prompts.
Fable gives you, what it learned. What people fancy as "infering your intent". Its slop
@MasterBayesian@thsottiaux apparently something good, as it keeps me winning right now with Astra 6.2, like doing complex financial modelling.
Something, Fable 5.2 did NOT manage to solve in that depth.
Why I’m not a doomer: it’s not that I doubt the capabilities explosion, actually.
It’s that the assumption that “more powerful destroys less powerful” is not some law of nature, it’s an imperial/colonial mind-virus that infects the people who happen to have a lot of social and economic power because they operate on those principles. This tends to skew towards a certain demographic.
They see the world as a question of domination, everything is a win/lose game. They can’t really believe, in their hearts, that other modes of relation exist.
So when I see people with high x-risk, I get it, but I just want them to like. Study ecologies and pre-imperial cultures. I’m not saying x-risk is zero, but there’s good reason to believe that sufficiently high capabilities will move beyond the mental trap that has been running western civilization since the Melian Dialogs.
Maybe I’m wrong. If I am, alright. That’s fine too, you know? But it’s kinda none of my business at that point, so why worry?
Meanwhile, valid critiques of the policies and practices that enable hypergrowth of AI capabilities at the cost of communities and economic security for ~everyone else mostly get ignored. Because the people most vocal there also have a demographic skew, and it’s not one that gets taken seriously by the media or the technorati or whatever.
But if I can see it you know Claude can see it, yk?
@MasterBayesian@thsottiaux Its the f'ing best model to talk to.
I havent sweared that little, since talking to ... not even Fable 5.2
Its a skill and task issue man
@BenjaminBadejo Exactly!!!
Its so hilarious that people are upset about, that the LLM does not just "infer" and therefore project the RL Pipeline onto the output, instead of being able to exactly specify what they want.
@tifafafafa@OpenAI@theo Like this comment said: "I'm convinced that Fable lovers under specify their prompts and Anthropic figured out that if they bake a little randomness into their output, people will mistake that as genius creativity."
Please dont repeat that shit, just to please @theo an friends
@tifafafafa@OpenAI its hilarious what @theo arguments and impression here are (like many others too): The model actually wants you to specify, while other models "infer" what you want to do..🤪
bro, if you let the model infer, you are not getting what YOU wanted, but what the RL Pipeline specified!
I'm convinced that Fable lovers under specify their prompts and Anthropic figured out that if they bake a little randomness into their output, people will mistake that as genius creativity.
Early testers tend to sling a lot of huge token burning tasks at models because they get a short window of infinite usage. This is not the typical experience of actually building something. Astra has proven AMAZING to me, but I tend to yap long instructions to build what I want with intention. And so when you just ask for "build me a new expansion in Majora's Mask" in Astra, you'll get a lukewarm return to the mean "Bellwhether" level.
To me, this is a symptom of OpenAI building tools vs Anthropic building a being. And it's sad to me that there was a strong day 1 preference for the model that rewards pure "pull the lever and see what comes out" style prompts (Fable.) That's technically cool, but it also represents total human replacement to me.
I prefer the tool to power me up instead of letting it take over. The voices I've seen discuss Astra as being KILLER for research, math, science, which are inherently more goal oriented. The kind of tasks that the user is driving rather than taking a ride.
Being real though I think we reached crossover months ago where you can use any frontier model to accomplish any task, including wild stuff outside of your domain of expertise. So use fable. Astra. Neon!!! They are all immaculate and complaining about anything today at the peak of history is pure wasted breath
@bdotdub Roll out Dots in EU, Germany.
Give reset to test new models.
Roll out Dots in EU, Germany.
Give reset to test new models.
Roll out Dots in EU, Germany.
Give reset to test new models.