@bryan_johnson Fund a biological institute to understand how aging works across lots of different animals. Do it as a side project. Do it at a marine lab (the have killer vibes). Do it for fun and interest. I 100% guarantee it will both be worth it and lots of fun.
While playing with Anthropic's demo of the Jacobian Lens, I found out that if you steer up the "FUCK" token in the 34th layer while asking Qwen 3.6 27B to lie, the model will just insult you
If you train a neural network to predict future words from past LLM neural activation representations. You can use that NN to find patterns in the brain of the LLM that maximally activate your NN for different words. You can then use those activations to probe the LLM for hidden thoughts or brain fuck it into thinking Paris is in Mexico.
@wholemars If they were really good they would license FSD. Collect all the driving data. Then distill their own FSD which they can optionally sell as a Tesla alternative.
Interesting, are you guys pairing ultrasound with mri to ground truth your data? This seems very ripe for machine learning. Doesn’t make sense to use ultrasound to predict mri data? Or is this pure physics based imaging. Costs would allow for very different kinds of study design. Would be very interesting for monitoring health of pregnant women.