AI has been going through hype cycles for about sixty years. I’m an old-ass software developer who’s been through enough hype cycles that I’ve lost my rose-tinted glasses. This will be the continuation of capitalism’s innovations in labor automation that’s been going on for centuries.
Every time I hear something like this I remember this phrase by Arthur C. Clarke which seems pretty good: “When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.”
Just because you can see all the ways in which it’s not going to work doesn’t mean that we can’t find a way around it.
And what kind of big discoveries were there on the last hype cycles that make it like what is happening now?
I’d say the Reflexion paper by itself already sets what is happening now apart from any other time, as it shows that inteligence, or something like it, can be achieved with what tech we have, even if it still needs some polishing.
Innovations in narrow AI will of course continue, and even accelerate in the short to medium term, given that the Eye of Venture Capital is fixed upon it right now.
But also, there is insane amount of hype, smoke, and mirrors about such innovations right now, and it is coming from the very same capitalist propaganda engine that hyped up blockchain/crypto/NFT/metaverse. It’s very convenient that all those graphics processors that were pumped out for crypto just happen to be transferable to neural networks. What more could the likes of NVIDIA & AMD ask for?
A lot of already insanely rich people in the tech sector stand to collect a whole bunch more by convincing dumb money (the wealthy people & institutions who fund VC firms) to invest in whatever new hype cycle they manufacture. Marc Andreessen of Andreessen Horowitz is not an honest person, he’s a P.T. Barnum for his VC firm. The people who just got kicked off the OpenAI board are of the same unhinged longtermism cult as Sam Bankman-Fried. Sam Altman has a net worth of half a billion; do you take his statements at face value?
If you want some dose of reality, I recommend the Marxist analysis podcast This Machine Kills, which focuses on technology trends. Every other episode is available for free from various podcast streaming outlets, like for instance Podtail.
But also, there is insane amount of hype, smoke, and mirrors about such innovations right now, and it is coming from the very same capitalist propaganda engine that hyped up blockchain/crypto/NFT/metaverse.
I don’t doubt it, I mean, Google themselves put out a presentation about their new AI a few weeks back that was fully faked, but there is still quite a lot of progress in the open source camp with ever better models appearing. I helped a little with one of the open source projects some months back and although that one might not have been the best thing it still had some results that were built upon. Anyway, open source is stil not to the point to threaten OpenAI but there are already many people using it for helping with work, for example.
It’s very convenient that all those graphics processors that were pumped out for crypto just happen to be transferable to neural networks.
Not that transferable from what I’m seeing especially as the best ones are needed for training new models while you don’t need as much to run them locally. What I have seen is more about new chips made for AI/Transformers, and in this case the newest and most advanced ones are being sold to the big tech companies at really high prices. I doubt they would be paying that much if they didn’t think it was needed, although to be fair, as long as it makes a profit it doesn’t need to actually be an advancement or anything good.
AI has been going through hype cycles for about sixty years. I’m an old-ass software developer who’s been through enough hype cycles that I’ve lost my rose-tinted glasses. This will be the continuation of capitalism’s innovations in labor automation that’s been going on for centuries.
Angela Collier: AI does not exist but it will ruin everything anyway
Every time I hear something like this I remember this phrase by Arthur C. Clarke which seems pretty good: “When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.”
Just because you can see all the ways in which it’s not going to work doesn’t mean that we can’t find a way around it.
Thanks, I’ve never heard this rebuttal before, except in every previous hype cycle.
But a fiction author said it! That means it must be true!
The web3 bubble cratered not ten seconds ago, but here we are again.
But this time my AI ape girlfriend will be able to love me back! (in 10 years after they complete the next firmware update)
And what kind of big discoveries were there on the last hype cycles that make it like what is happening now?
I’d say the Reflexion paper by itself already sets what is happening now apart from any other time, as it shows that inteligence, or something like it, can be achieved with what tech we have, even if it still needs some polishing.
I’m not sure what you’re referring to. Maybe this: “Reflexion: Language Agents with Verbal Reinforcement Learning.” Whatever you meant, I haven’t read it.
Innovations in narrow AI will of course continue, and even accelerate in the short to medium term, given that the Eye of Venture Capital is fixed upon it right now.
But also, there is insane amount of hype, smoke, and mirrors about such innovations right now, and it is coming from the very same capitalist propaganda engine that hyped up blockchain/crypto/NFT/metaverse. It’s very convenient that all those graphics processors that were pumped out for crypto just happen to be transferable to neural networks. What more could the likes of NVIDIA & AMD ask for?
A lot of already insanely rich people in the tech sector stand to collect a whole bunch more by convincing dumb money (the wealthy people & institutions who fund VC firms) to invest in whatever new hype cycle they manufacture. Marc Andreessen of Andreessen Horowitz is not an honest person, he’s a P.T. Barnum for his VC firm. The people who just got kicked off the OpenAI board are of the same unhinged longtermism cult as Sam Bankman-Fried. Sam Altman has a net worth of half a billion; do you take his statements at face value?
If you want some dose of reality, I recommend the Marxist analysis podcast This Machine Kills, which focuses on technology trends. Every other episode is available for free from various podcast streaming outlets, like for instance Podtail.
I don’t doubt it, I mean, Google themselves put out a presentation about their new AI a few weeks back that was fully faked, but there is still quite a lot of progress in the open source camp with ever better models appearing. I helped a little with one of the open source projects some months back and although that one might not have been the best thing it still had some results that were built upon. Anyway, open source is stil not to the point to threaten OpenAI but there are already many people using it for helping with work, for example.
Not that transferable from what I’m seeing especially as the best ones are needed for training new models while you don’t need as much to run them locally. What I have seen is more about new chips made for AI/Transformers, and in this case the newest and most advanced ones are being sold to the big tech companies at really high prices. I doubt they would be paying that much if they didn’t think it was needed, although to be fair, as long as it makes a profit it doesn’t need to actually be an advancement or anything good.