Kind of both, we're still on an exponential curve if you zoom out but its's like a giant exponential made up of many sigmoid curves. Currently we're in a bit of a trough of disillusionment after a massive amount of hype. Perhaps the language models that are currently being trained will blow our minds when they come out in a few moths. I suspect they'll move up from intelligent high schooler that has access to an internets worth of knowledge but has poor long-horizon reasoning and robustness to a PHD level of intelligence with much better logic, robustness, and longer term reasoning. That might be a little disappointing, or it might start another hype cycle. Whenever there's hype, there's more money, more algorithmic breakthroughs, more compute etc. Long story short the world isn't going to totally transform in a year, but by the 2030s we'll look back and go "wtf". Humans adapt really quickly though. Buy a fast car and it'll be fun for a month, then it'll just be another car and you'll want an even faster one. Same thing.
Fantastic. But please can someone carefully edit the transcript, which is often frankly misleading, also explaining the many obscure acronyms bandied about! Thanks!
A global centralised consortium on AI == Pure fantasy at this time. The most likely version of the consortium hypothesis would be multiple consortiums formed along broad political/value/power based fault lines, which would not achieve the stability/safety objective of the consortium hypothesis.
A multinational AGI consortium (MAGIC, look it up on arxiv) might incentive countries to join the treaty in an attempt to get a stake in the AGI pie, leveraging their self-interest for global coordination.
A bit too much Pinker-ism for my taste. The AI revolution will further entrench historic inequality in the US, and further empower the 1% that effectively already controls most of governance. This revolution will not pan out well for most people in the US and those on the rest of the planet, at least for the next generation. But the long termer aspect of e/acc philosophy as I understand it would just scoff at those social justice concerns and make the long term argument that what happens in 100 years as a consequence of AI accelerationism is the most important. As if they can accurately access that to begin with. They can't. Meanwhile, unregulated deep fake platforms super charge the powers of scammers. But to that? Meh! Equating "both sides" regarding what is happening in the US just illustrates how clueless some people can be to the rise of fascism, and which side is most responsible for the situation of the rise of threats of political violence and political violence. There is NO comparison in that regard. Not even close.
REBUTTAL: Even without AI the 1% will become further and further entrenched. Consider the hypothetical scenario without AI... what can the 99% and future generations look forward to? I see a massive potential for increasing inequality and marginalisation with reprieve that comes at a rate to slow to outpace the extraction of labor created by this inequality. The technological clock turns forward while the clock for equality and life satisfaction and fulfilment begins to unwind as every free moment of the proletariat is shoved into maintaining subsistence as technological advancement can't keep pace with the extraction rates of labor demanded by the 1%. With AI there is at least hope that we can have the technological progress to end scarcity and give humanity a completely different trajectory to the one described above or anything remotely close to it. How else do we free ourselves from the shackles of capitalist governed greed where the bourgeoisie dictate the the outcomes of the proletariat unless we remove the actual need for the proletariat as a labor force entirely without the requirement of revolution and war (which would be entirely futile at best given the non-localised state of wealth distribution). REBUTTAL2: Equating "both sides" Both sides have their flaws. Both sides cannot and should not be summarised by the extremists who exist on both ends. The fact that I have no idea which side in particular you belong to (and you clearly belong to one by the tone of your comment) shows that your comment cuts both ways. It could be equally applied to the left or the right, as much as you might assert it belongs only to whichever is the opposite side to yours (and I literally cannot tell which side you're on by your comment). I see fascists on the left and I see fascists on the right. I see members from both sides who claim that their ideological purity gives them the right to silence, sanction and punish those with dissenting opinions. The very definition of fascism. I see leftists who want to silence anyone for daring utter words that offend anyone who, in the name of freedom, are against free speech, to protect the "free" people from the unknown horrors and chaos of free speech. I see leftists who are intolerant of intolerance and willing to persecute anyone who acts with what they define as intolerance : conveniently ignoring the fact that it requires some governing body to decide what is considered intolerant. Which governing body? Well, them of course. It's literally Animal Farm. All animals are equal but some animals are more equal than others. I see rightists who want to impose their conservative beliefs on everyone. They want to control the wombs of women. They want to instigate the teaching of religion in schools under the guise of "intelligent design" and call it science. They want to force LGBTs back into the closet so hard that they never come out again. They would if they could make it illegal to be openly LGBT. The real problem isn't AI, the problem is human bipolarisation, and the worst actors from both sides are already working in full force with existing algorithmic technology to drive a permanent wedge between these ever widening groups. I really don't see AI making this any worse, and this is literally what the 1% want. Go read about Noam Chomsky 10 strategies of manipulation. The anger for the "other" is the fuel that keeps us from ever uniting to actually do anything about the 1%.
Nah I liked hearing vitaliks take on everything we’ve heard before. I guess it’s kind of a nothing burger but only because a lot of intellectuals agree on a lot of it
So are we finally on the upward curve or is this just another springboard bounce before the actual leap?
Kind of both, we're still on an exponential curve if you zoom out but its's like a giant exponential made up of many sigmoid curves. Currently we're in a bit of a trough of disillusionment after a massive amount of hype. Perhaps the language models that are currently being trained will blow our minds when they come out in a few moths. I suspect they'll move up from intelligent high schooler that has access to an internets worth of knowledge but has poor long-horizon reasoning and robustness to a PHD level of intelligence with much better logic, robustness, and longer term reasoning. That might be a little disappointing, or it might start another hype cycle. Whenever there's hype, there's more money, more algorithmic breakthroughs, more compute etc. Long story short the world isn't going to totally transform in a year, but by the 2030s we'll look back and go "wtf". Humans adapt really quickly though. Buy a fast car and it'll be fun for a month, then it'll just be another car and you'll want an even faster one. Same thing.
Fantastic. But please can someone carefully edit the transcript, which is often frankly misleading, also explaining the many obscure acronyms bandied about! Thanks!
*IT'S ALL COORDINATION*
3 hours of bliss
Havent heard from him in a while
Nice
My brain just expanded
A global centralised consortium on AI == Pure fantasy at this time.
The most likely version of the consortium hypothesis would be multiple consortiums formed along broad political/value/power based fault lines, which would not achieve the stability/safety objective of the consortium hypothesis.
A multinational AGI consortium (MAGIC, look it up on arxiv) might incentive countries to join the treaty in an attempt to get a stake in the AGI pie, leveraging their self-interest for global coordination.
80K DON'T STOP
So you'd board a plane with a 10% chance of crashing on your next trip? Not that there is any reasonable way to stop what's coming.
Truuuuuuuu
epic
58:52
1:34:59
A bit too much Pinker-ism for my taste. The AI revolution will further entrench historic inequality in the US, and further empower the 1% that effectively already controls most of governance. This revolution will not pan out well for most people in the US and those on the rest of the planet, at least for the next generation. But the long termer aspect of e/acc philosophy as I understand it would just scoff at those social justice concerns and make the long term argument that what happens in 100 years as a consequence of AI accelerationism is the most important. As if they can accurately access that to begin with. They can't. Meanwhile, unregulated deep fake platforms super charge the powers of scammers. But to that? Meh!
Equating "both sides" regarding what is happening in the US just illustrates how clueless some people can be to the rise of fascism, and which side is most responsible for the situation of the rise of threats of political violence and political violence. There is NO comparison in that regard. Not even close.
REBUTTAL:
Even without AI the 1% will become further and further entrenched. Consider the hypothetical scenario without AI... what can the 99% and future generations look forward to? I see a massive potential for increasing inequality and marginalisation with reprieve that comes at a rate to slow to outpace the extraction of labor created by this inequality. The technological clock turns forward while the clock for equality and life satisfaction and fulfilment begins to unwind as every free moment of the proletariat is shoved into maintaining subsistence as technological advancement can't keep pace with the extraction rates of labor demanded by the 1%.
With AI there is at least hope that we can have the technological progress to end scarcity and give humanity a completely different trajectory to the one described above or anything remotely close to it. How else do we free ourselves from the shackles of capitalist governed greed where the bourgeoisie dictate the the outcomes of the proletariat unless we remove the actual need for the proletariat as a labor force entirely without the requirement of revolution and war (which would be entirely futile at best given the non-localised state of wealth distribution).
REBUTTAL2: Equating "both sides"
Both sides have their flaws. Both sides cannot and should not be summarised by the extremists who exist on both ends. The fact that I have no idea which side in particular you belong to (and you clearly belong to one by the tone of your comment) shows that your comment cuts both ways. It could be equally applied to the left or the right, as much as you might assert it belongs only to whichever is the opposite side to yours (and I literally cannot tell which side you're on by your comment).
I see fascists on the left and I see fascists on the right. I see members from both sides who claim that their ideological purity gives them the right to silence, sanction and punish those with dissenting opinions. The very definition of fascism.
I see leftists who want to silence anyone for daring utter words that offend anyone who, in the name of freedom, are against free speech, to protect the "free" people from the unknown horrors and chaos of free speech. I see leftists who are intolerant of intolerance and willing to persecute anyone who acts with what they define as intolerance : conveniently ignoring the fact that it requires some governing body to decide what is considered intolerant. Which governing body? Well, them of course. It's literally Animal Farm. All animals are equal but some animals are more equal than others.
I see rightists who want to impose their conservative beliefs on everyone. They want to control the wombs of women. They want to instigate the teaching of religion in schools under the guise of "intelligent design" and call it science. They want to force LGBTs back into the closet so hard that they never come out again. They would if they could make it illegal to be openly LGBT.
The real problem isn't AI, the problem is human bipolarisation, and the worst actors from both sides are already working in full force with existing algorithmic technology to drive a permanent wedge between these ever widening groups. I really don't see AI making this any worse, and this is literally what the 1% want. Go read about Noam Chomsky 10 strategies of manipulation. The anger for the "other" is the fuel that keeps us from ever uniting to actually do anything about the 1%.
Hey I love making videos. Would love to work with you!
I love memes
3 часа!... издеваетесь?) Да ещё и на аГлицком😮
Nothing 🍔
Nah I liked hearing vitaliks take on everything we’ve heard before. I guess it’s kind of a nothing burger but only because a lot of intellectuals agree on a lot of it
@@EthereumProtocol
😁👍