Sorry for the vague title. I'm avoiding spoilers for those who care a lot about that sort of thing.


The series illustrates a fast take-off scenario. It tells the story of a fast-take-off scenario of "uploaded intelligence". Of course the usual fast take-off scenario this community is concerned about is artificial intelligence, not uploaded intelligence (i.e., digitized human intelligence as in Robin Hanson's Age of Em). But the series portrays the dangers of uploaded intelligence in a remarkably similar way to the concerns many people have about a fast AI take-off.

It feels very embedded in the state of the current world in 2022 in many ways; referencing the pandemic in ways most shows don't dare to; set in perhaps an alternative history where we weren't on the cusp of AGI, but where a Steve Jobs-like character had almost, but not quite, cracked the brain-uploading problem 20 years ago.

There are plenty of plot devices that don't feel realistic, but the interaction between technology and geopolitics, the exponential growth when AI take-off happens, the pivotal nature of very specific events guiding the direction of the take-off at the time it occurs--all that tracks, for me.

Worth watching as entertainment for yourself, and worth popularizing as a way to prime people out there on the dangers of a fast AI take-off in the near future.

That is all I have to say! Just putting the alert out there.

Sorted by Click to highlight new comments since: Today at 10:52 AM

(it'd be handy to have a link in the opening paragraph so if I wanna avoid spoilers I can go do that easily)

Ben - looks like an interesting series, and it scores pretty well on IMDB so far (8.1/10, but with a fairly low sample size of 1,000). 

I do find it frustrating though that most fictionalized portrayals of AI development and impacts over-use the same old tropes to create dramatic tension and conflict, e.g. the 'evil corporation', the 'plucky young friends fighting for humanity', 'but what about human emotions and relationships', etc.

It would be really helpful, I think, for AI safety/EA people to develop an explicit critique of how AI developments are dramatized by lazy screenwriters, and how that leads the public to misunderstand AI risks.

I agree. I'm working on a project now to understand cognitive biases which might mislead the public on AI risk and perhaps media studies is another area to explore those misunderstandings. 

I will say that in Pantheon, the "evil corporation" trope feels like it gets a bit more nuanced as the series reaches its last few episodes.

Ben -- let me know if you'd like to chat further about your project; sounds very interesting and useful. I've been studying cognitive biases on and off for the last 30 years. (Best way to reach me is by email, which is on my site here.