@bryan_johnson Hmm…. I wonder why agents would chose, of all possible behaviors, the same as common forum degens 🤔 and I that thought religious organization was a human thing, separate from the information content.
@BadreNicolas Awsome thread! The representation of schizophrenia is indeed heterogeneous and varied greatly over time. Would be possible to dissect how F/SGAs perform on the core phenomenological symptoms, like aberrant salience?
@saxena_puru Just waiting to see what happens when AI agents are mass deployed in the hands elite AI enthusiasts - a million more home made quant bots. What could possibly go wrong?
@Paddy_Barrett, thinking about zone cardio training, what’s your take on running with beta blocker?
Lowering heart rate zone while maintaining peripheral demand, EPO, Hb, vascularization etc.
Conversely, for pts on ADHD meds - higher zone at lower effort.
No one that’s reacting to that ALL of the DeepSeek benchmarks are VERY close to other models?
Something that comes from zero would probably have a much more unbalanced profile.
I doubt this would have happened before any of the other awesome open source models were out.
People question LLMs ability to reason, as it is merely prompting itself in iteration.
But really, how much do the average person reason before the average decision?
That (moral) would be more like a overfittet simple regression.
I would love dr @JonHaidt take on this!
@Giuliano_Mana Behave and Determined by Robert Sapolsky! They are mind blowing.
Also, Righteous mind by Haidt will change how you think about knowledge