Who wants to test this with a practical experiment? Ill pay the Claude/GPT4 bill

so much to do

I’m an e/acc but this take is incredibly ignorant

We built pseudo intelligent wordcels that can spit out infinite synthetic semi-garbage BUT with the right verifier they can rotate shapes so study verifiers

'By combining the power of synthetic data with a rigorous verification process, AlphaGeometry achieves a problem-solving ability comparable to that of a human Olympiad gold medalist...'

'Rigorous' verifiers are the final frontier of AI where are they

The point of the universe

Big update:

- We partnered up with @openstore
- The AI-generated ads are performing 3x better than baseline 🤯

Compute as the bottleneck is a fundraising meme

less GPUs predicting next token more GPUs predicting the future

kinda ate

As above, so below

Another amazing biohackers meetup in NYC yesterday.

Main topic: exploration of unreasonably effective compounds

Recap:

- Discussed effectiveness of Soviet era peptides (e.g. Semax, Selank)

- Soviets were on a tear in the early 80’s, discovering game changing compounds. How

the thing about peptides is they actually work

What do we think Joscha Bach

oh? you're not a biohacker?

cool, just let the neuro-exploitative capitalistic meta-entity rewire your ancestral code for profit

sick, no that's great

Anthropic is so in it’s not even funny

We need a word to describe people who think the upper bound of human health is achievable by just sleeping and working out

Naturalistic fallacy is rampant rn

a fund that invests in crypto to fund AI to fund longevity research