Conscious AI is the second scariest kind

DishBrain, described in this fantastic article, apparently figured out how to play Pong [better than `chance`] without being told to "achieve a high score" — using petri-dish-grown brain cells with only a very basic input/output [limbic] system.

Basic info on DishBrain: https://www.perplexity.ai/search/What-can-you-GK9aohldRJqh9o...

----

I love Author Watts' comparison that grasshoppers have less neurons than DishBrain, and yet are often considered sentient.

----

My only [potential?] "fact check" would be that the inventor of the Paper Clip Maximization Theory is actually Ilya Sudkowski.

An interesting article by Peter Watts, author of Blindsight, on why AI consciousness may not be the most scary thing that could emerge.

A great piece of writing.

Text-only, no Javascript, works where archive.is is blocked:

    curl -si40A "" https://www.theatlantic.com/ideas/archive/2024/03/ai-consciousness-science-fiction/677659/ \
    |grep -o "ArticleParagraph_root.*</p>" \
    |sed 's/.\{61\}/<meta charset=utf-8\/>/' > 1.htm

    firefox ./1.htm
    links -no-connect -force-html 1.htm

Disabling Javascript also works. Ads require Javascript, reading the article does not