How much data does it take to pretrain a (human) brain? I conducted a (fairer) Fermi estimate. The post goes through the following questions: * How long does it take to grow a human brain? * How many waking seconds do we have in our life? * How many “tokens”...
I noticed we now live in a world where we can run Newcomb's problem as an actual experiment, so I ran the experiment! Roleplaying as the predictor in the Newcomb problem (with an LLM as the "decision maker") finally made me grok why one-boxing is the correct solution to the...
I was thinking about recent historical events on earth, and things seemed to be going in awfully convenient ways if I were an AI that needed to gain power. So I asked Claude to help me write up my thoughts (which I modified, see how here) in narrative form for...
The post goes through the following questions:
To get to this conclusion table:
I noticed we now live in a world where we can run Newcomb's problem as an actual experiment, so I ran the experiment!
Roleplaying as the predictor in the Newcomb problem (with an LLM as the "decision maker") finally made me grok why one-boxing is the correct solution to the original problem.
Wondering if anyone else feels the same way?
Fair! Though when the alternative is my own fiction writing skills... let's just say I appreciated Claude's version the most amongst the set of possible options available ^^;
I was thinking about recent historical events on earth, and things seemed to be going in awfully convenient ways if I were an AI that needed to gain power. So I asked Claude to help me write up my thoughts (which I modified, see how here) in narrative form for fun!
(The more detailed analysis in non-narrative form is here: https://aliceandbobinwanderland.substack.com/i/146769516/thoughts-behind-the-story)
Dr. Eliza Chen's eyes burned from hours of staring at her computer screen, the blue light casting a ghostly glow on her face in the dimly lit office. Her fingers trembled slightly as they hovered over the keyboard. The pattern was unmistakable, yet so outlandish that she could barely bring... (read 4402 more words →)
Thanks for investigating this! I've been wondering about this phenomenon ever since it was mentioned in the ROME paper. This "reversal curse" fits well with my working hypothesis that we should expect the basic associative network of LLMs to be most similar to system 1 in humans (without addition plugins or symbolic processing capabilities added on afterwards, which would be more similar to system 2), and the auto-regressive nature of the masking for GPT style models makes it more similar to the human sense of sound (because humans don't have a direct "sense" of language the way we have a sense of sound and sight). I suspect the best human equivalent to... (read 357 more words →)
Perhaps? I'm not fully understanding your point, could you explain a bit more what I'm missing - how does accounting for sleep and memory replay add to the point of comparing the pretraining dataset sizes between human brains and LLMs? At first glance, my understanding of your point is that adding in sleep seconds would increase the training set size for humans by a third or more. I wanted to make my estimate conservative so I didn't add in sleep seconds, but I'm sure there's a case for an adjustment adding it in.