Macademe pfp

Macademe

@smokey-pete

204 Following
39 Followers
Macademe pfp
Macademe
@smokey-pete
What are you going to Venice for? Biennale?
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
No eating past 7 pm 3 fruits a day
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
Do relaxation exercises once a day (eg guided hypnosis or slow breathing or tai chi)
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
What does 100k refer to here? Price?
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
@s-r-constantin
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
https://twitter.com/s_r_constantin/status/1802737462776676716
1 reply
0 recast
1 reaction
Macademe pfp
Macademe
@smokey-pete
Reminds me of leashless saying that the US army should run the protection of the Amazon rainforest as planetary infrastructure for o2 generation
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I don’t understand why you call this hardening, to me that sounds more like „investing in“ or „growing“ the commons. To me hardening has connotations of defense against attack, which eg planting trees doesn’t do.
1 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I thought you might appreciate this demonstration of decoding straussian messages via ChatGPT https://x.com/eigenrobot/status/1792748484845805672?s=46 With the conclusion that only one straussian reading exists 😂
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
What’s her handle? I love this
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
Amazing video, I wasn’t expecting it to be that good!
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I like the Taptic Engine bumps when the chatgpt app generates text on iOS. So satisfying!
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
Apologies for handwaving here, we’re not really solving for these distributions, just generating samples but maybe there’s still something there
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
So the real innovation is learning the prior I think. I’m not going anywhere in particular with this but I think you’re right. I guess next level is conditioning p(X) in ever more sophisticated ways to control generation
1 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
True, one is a forward problem and the other is an inverse problem. Predicting p(X|y) is harder than the reverse if X is high dimensional. To make this possible we came up with a way to learn a very good prior on X using big NNs. The prior constrains the soln of inverse problem.
1 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
Can you say or point to a post about what are the circumstances that have shifted? Geographic distribution of talent sounds like one, maybe. Something about the internet but I can’t figure out exactly what becomes different. Because there’s still huge value in hanging out IRL with talented people.
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I mean I think it’s straightforwardly supposed to be the socium, eg dictatorship of the proletariat
1 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I need a hobby. Don’t like being bad at new stuff tho
0 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I really enjoyed this article and was happy to see the shop is still around as well. However, why gloom?
1 reply
0 recast
0 reaction
Macademe pfp
Macademe
@smokey-pete
I also wonder if this is my case
0 reply
0 recast
0 reaction