Rippling Co-founder tweets notes of convo with Sama then deletes
twitter.comAll these tech "influencers" and CEOs have such laughable opinions on AGI. First, they cannot really seem to give a formal definition of AGI (are we talking about some network that can problem solving in n > 2 problems or defining and creating consciousness)? Second, it's clear many of these people are just bandwagoning onto the hype. They started with crypto, then maybe some VR, and now are moving to AI/AGI coat-tailing on the accomplishments of OpenAI and DeepMind. What exactly about AGI do they find compelling or find feasible? Are they asserting that a deep learning networks of 2 separate problem spaces can be combined in an elegant manner? It's all just spew that looks like some article that can be posted on Vixra.
These tech "influencers" and "big CEOs" know no shame in their constant attention seeking. I guess to be fair, many people on HN or other platforms talk about AGI and other popular things, but the statements and arguments generally are a lot more specific or nuanced, relative to these quackery beliefs from "influencers" who are clearly attempting to gather a digital "I am smart" badge. Just lol.
> it's clear many of these people are just bandwagoning onto the hype. They started with crypto, then maybe some VR, and now are moving to AI/AGI coat-tailing on the accomplishments of OpenAI and DeepMind.
I'm impressed with how they've kludged crypto hype into the AI-hype discussion actually. It's going to provide "civilizational equilibrium" no less, because AGI is apparently just buying compute time, which obviously can't be done with money.
And I guess we need to buy into their decentralized network so that this AGI compute can be done on their "clearly unbiased" crypto software that's running on the computers of all those who bought into the hype of their crypto coin. And then every participant who worships the feet of the crypto lords can get a unique NFT of their crypto lord. Maybe one person will get an NFT with the crypto CEO having silly green glasses, or another person who helped the CEO become a billionaire will get an NFT with the CEO having a pink mohawk. The endless possibilities of crypto and AGI :-)
It's certainly a potential path. It's folding at home model + crypto. May offset electricity costs if you can sell the tokens...
I am interested in understanding why more training flops vs say more parameters. Surely bigger models perform better but perhaps there is a limit?
I don't quite understand how distributed subtrees will work.
I'm not an expert in ML theory, but from my perspective, I think more compute would cause better fitting for a model, but possibly overfitting if more parameters are not added. For some problem space, adding additional parameters introduces another degree of freedom, which should allow a larger domain for the inputs mapping to outputs (more answers to questions). And if we define AGI as a network that can answer questions for n > 2 domains (e.g. can we do image classification and a chat bot and synthesize them into a coherent system passes a Turing test), then more parameters makes sense to increase the range of outputs.
Interestingly, I don't think it's clear on how parameters in the network and compute would clearly find a domain within a combined problem space, where the mapping from question to answer will give sensible results. It seems like we need more tools to extend ML.
We know more parameters improves generalization so that is why it's unclear to me what they mean by more training flops.
One new parameter = 1 new node on the crpto nutter's network to do additional compute for the parameter? Lol I can see many ways these crypto nutters can attempt to weakly justify purchasing into their network.
While what you are saying is mostly true for most of the "influencers", this guy was a 2x icpc world finalist back in the day, so probably knows a thing or two. Having said that, I read through the excerpt and it is actually outright wrong on some stuff.
Makes sense. While he certainly seems to be smart given his accomplishments, I think it's really up to the general educated populace to keep those who are in power and have technical understanding of things in check. It seems clear that there's an ever growing of "influencers" who do have technical skills, but are saying largely quixotic and exaggerated statements to boost their own self image or products.
Except sama has quite explicitly said that these LLMs are not AGI and that there's a long way to go. Presumably it was deleted because either: it is counter to the lie sama is telling about his belief about what is/isn't AGI... or the author was mistaken in their understanding of what sama said. I assume the latter, since ChatGPT is obviously not AGI.
"[...] interesting to me how many of the ChatGPT takes are either "this is AGI" (obviously not close, lol) [...]"
I wish some of these ‘big names’ would make an effort to define all the terms. Which people are not into hyping but could be part of defining what AGI is, so we can put this ‘we have it in a couple of months’ to bed?
Carmack is doing an AGI startup; maybe he can offer a definition of what it is exactly his company will attempt to deliver and how we can recognise it when/if it does?
If there's any debate on whether the Dunning–Kruger effect is real, just look at a 'tech influencers' tweets about neural networks.