Account · 2 tweets

spicylemonade

@spicey_lemonade ↗

Tweets

spicylemonade @spicey_lemonade
Affirming

Math, of course, cannot be conscious. Let’s assume a human could perfectly write/simulate every output of a 100T parameter LLM on a piece of paper. That system, in itself, is conscious. That system would have to be aware that a human is writing the outputs for it when you ask it.

Ted Lieu @tedlieu

I like the below joke. Another version: math equations are not conscious. Rocks are not conscious. If you attach the above to random number generators, they still wouldn’t be conscious. We don’t need to define consciousness to know the Pythagorean theorem is not conscious.

retweets 4 likes 61 quoted by 1 archive account substance 6 · argument · rank 186 · A system simulating a massive LLM, even on paper, would be conscious. view on X ↗
1 top reply
Aleksi Liimatainen @aleksil79

query the llm in chinese and you have exactly replicated searle's chinese room thought experiment. it's been around since the 1980s and you can see how much that's helped

spicylemonade @spicey_lemonade
Affirming

You can calculate how a clump of neurons will react to electrical impulses by hand. It’s a math problem. Clearly a large clump of neurons can’t be conscious!!

retweets 0 likes 1.3k substance 6 · argument · rank 90 · Calculable physical systems can still be conscious, refuting the math-based argument against AI consciousness. view on X ↗
1 top reply
MrSane @MrSaneApps

Turns out you cannot “Perfect prediction of any brain state demands complete knowledge of every relevant particle, synapse and fluctuation—blocked in practice by chaos, incomplete measurement and possible quantum effects. LLMs are discrete deterministic maps from known weights