Dr. Mike Israetel

-PhD in Sport Physiology and former Prof. of Nutrition and Sport Science
-Cofounder of Renaissance Periodization
-BJJ Black Belt, Competitive Bodybuilder

[Dr. Mike Israetel @misraetel]

@WesRoth 🤣

[Dr. Mike Israetel @misraetel]

@tszzl 100%

[Dr. Mike Israetel @misraetel]

@creatine_cycle 🙏🙏🙏🙏

[Dr. Mike Israetel @misraetel]

@coleywoleyyy @dwarkesh_sp @_sholtodouglas @mintlify Shoot me a DM!

[Dr. Mike Israetel @misraetel]

Huge fun training the @dwarkesh_sp podcast folks as well as a lone, wandering @_sholtodouglas. These guys CRUSHED! And only one person died! Success!

[Dr. Mike Israetel @misraetel]

@abhinavRV @dwarkesh_sp @_sholtodouglas It was me!!!! Please say hi next time or at least throw a rock at me!

[Dr. Mike Israetel @misraetel]

@_sholtodouglas @dwarkesh_sp 🙏

[Dr. Mike Israetel @misraetel]

@Dr_Singularity 100%

[Dr. Mike Israetel @misraetel]

@dwarkesh_sp @_sholtodouglas You’re the man!!!

[Dr. Mike Israetel @misraetel]

@cremieuxrecueil 🙏🙏🙏

[Dr. Mike Israetel @misraetel]

Check it out!

[atlas @creatine_cycle]

i took @misraetel to the park to see the dogs and talk about classic san francisco topics such as:
- the permanent underclass
- artificial intelligence
- human enhancement and peptides

everybody has been waiting to hear what the bodybuilders have to say about all this.

00:00 - San Francisco and the permanent underclass
06:00 - Is the permanent underclass discussed outside of SF?
10:35 - Mike’s AI timelines and the future of work
19:50 - Timelines on robot workers
24:20 - The three phases of AI progress
32:20 - Dogs and Ads
34:57 - UBI and when society won’t have to worry about capital
38:58 - Humans won’t need much in the singularity
42:55 - Human self-improvement is the next major category
52:05 - @MartinShkreli shoutout and the future of medicine
1:00:19 - @beffjezos shoutout and why machines won’t turn on us
1:09:35 - AI alignment is important and the Mythos moment

[Dr. Mike Israetel @misraetel]

Attractive people often get more patience, more benefit of the doubt, and more chances. It’s not fair, but it is real.
https://t.co/ndcZGIW9qP

[Dr. Mike Israetel @misraetel]

@creatine_cycle Bunch of sperm racing LOSERS if you ask me?

Can’t hang? Don’t jizz! Hate the game, not the spermer.

[Dr. Mike Israetel @misraetel]

@RobertDMellish I love this question. To me the best answer is: we have done, as a civilization, so little on the grand scale that the amount left to do still is gonna need all hands on deck, machine and human.

[Dr. Mike Israetel @misraetel]

What do you guys think of my reply?

[Dr. Mike Israetel @misraetel]

@scaling01 I appreciate the critiques, but I think we're probably not so far off on our opinions as this clip might suggest. A few things:

1.) "continues to only talk about 4 billion robots that are equivalent to 4 billion humans"

On the one hand, robots will be, at 4 billion units of humanoids out and deployed, way more productive than humans in very many tasks, absolutely.

On the other hand, if we say that's the functional equivalent of 16 billion humans, it's still not clear to me that this can saturate ALL of human labor demand. Compared to ancestral times where we had 100 other workers in a tribe to trade with, our modern multi-billion-person economy has still yet to saturate all human demand. It's just not clear to me that anything in the single digit multiples of labor from today can do this. I think it's easy to envision every human having at least 3-5 robots, and that doesn't even count enterprise robot fleets, which would number way higher per person.

2.) "so he completely misses the fact that robot population would grow exponentially at 2-10x a year, implying that we would cross his magical threshold of 10x more robots than humans within at most 3 years"

I actually attend to this at the very end of the clip, with the claim that 10x the human labor force might start to supply a very large fraction of human demand. And not only do I agree with you here, I've been saying this sort of thing for years on my channel. But, the good news is that once we have 10x+ the robots we do humans, the wealth they generate makes human employment or otherwise largely irrelevant to human standards of living.

3.) "also misses the fact that AI will be already be much more intelligent than humans by the point we have 4 billion robots doing real economically valuable work"

I've made this prediction dozens of times and I agree with you. I suspect that by 4 billion humanoids, AI in the cloud will be 100x or 1000x+ more intelligent than humans. I think this high a level of intelligence will be able to deploy humans way more valuably than we've been deploying ourselves. The goal at this point is no longer to just satisfy human wants, but to build a galaxy-wide civilization. Almost any additional help is going to be valuable; human or otherwise.

4.) "AI 2040 scenario predicts 200 BILLION human equivalent AI workers by 2036"

I think with computing advances it might even be substantially bigger than this. But, the number of physical robots in the world might be much smaller. I actually think by 2040 shit will get really crazy with robot numbers, but we have to keep in mind that until very high robot numbers, computer-only AI instances will outnumber robot workers (and human workers) by such massive margins that the net productivity of society will be greatly limited by the number of physical hands in the real world to do work. For this reason, I don't think humans will be out of work until well into orders of magnitude more robots than humans deployed.

[Dr. Mike Israetel @misraetel]

@Talionbrimbor @scaling01 Because so far, orders of magnitude more people than historically have flooded the market and humans still care about work!

[Dr. Mike Israetel @misraetel]

@scaling01 Maybe! But I think the machines will have many more productive uses for us than we have for ourselves for longer than many think.

[Dr. Mike Israetel @misraetel]

"METR-style task reliability is doubling every 40 days, on track for one model to outwork every human who ever lived by 2032, while Benedict Evans argues tokens are destined to be low-margin commodity infrastructure. Superintelligence too cheap to meter, arriving on schedule."

[Dr. Alex Wissner-Gross @alexwg]

https://t.co/Ptd78653wi