Mati_Roy's Comments

How much delay do you generally have between having a good new idea and sharing that idea publicly online?

Mine is probably much longer than it should. Although I also have some reasons not to share them right away such as needing them to have a good first impression.

I probably haven't shared most of them yet, so the delay is probably at least a few years currently.

Distinguishing definitions of takeoff

trying to put this in my own words to remember it

so different axes for take-off dynamics include:

  • time span: physical time, economic time, political time, AI time (development speed of front runners over others)
  • shape of the take-off curve, ex.: exponential, S-curves, linear, etc.
  • monopolistic effect: do front runners become less likely to be outcompeted as they grow? how many large players will there be? also: how will this change? ex.: it could be that AI doesn't have strong monopolistic effect until you reach a certain level

measurement to quantify:

  • AI progress: GDP, decisive strategic advantage
  • AI progress speed: time until AI having more power than the rest of humanity / time until solving the control problem


  • impact on forecasting capabilities
Distinguishing definitions of takeoff

the fraction of global economic activity attributable to autonomous AI systems will rise

I thought a bit about this, but haven't figured it out: how can this be measured? if AI is commoditized, AI companies won't make a profit from it. AI researchers might make more money, but likely not more than however much it would cost to train more AI researchers (or something like that). maybe we can see which industries have their price reduced because of AI, and count this as a lower bound for the consumer surplus created by AI. what else?

What are the risks of having your genome publicly available?

I personally would rather an FAI be able to bring me back than preventing an UFAI from doing so

The housekeeper

when we were more people at the Macroscope (, the last tool we were using was weighted reverse auctions:

What plausible beliefs do you think could likely get someone diagnosed with a mental illness by a psychiatrist?

Thanks for the clarifying question. It's the impersonal "you"; although I think specifically the personal "you" would be even more interesting. I updated the question.

What plausible beliefs do you think could likely get someone diagnosed with a mental illness by a psychiatrist?

Believing you're likely in a simulation

Various lifestyles: believing that polyamory can be an healthy relationship structure; believing practicing various kinks and sexual interests can be healthy

What are beliefs you wouldn't want (or would feel apprehensive about being) public if you had (or have) them?

Thank you for your answer! I'd want you to expand on all points, but will be more reasonable and prioritize my curiosity; re 2: does it creates more failures than successes because a) more people start a startup as a result, thinking it will be easier than it is, b) harder to get profit because competition is more fierce, or c) other reasons?

Whole Brain Emulation: Looking At Progress On C. elgans

For your information, the above two links were judged as wrong

@davidad, any updates on your work?

What could a World Unification Index track to measure how unified the world is, was, and is becoming?

I have the impression that the more unified the world is, a) the less likely wars are, b) the easier it is to coordinate to solve global problems. It would also create more economic prosperity. But some forms of globalisation could also make our civilisation less robust. Anders Sandberg says the extent a feature is globalize should be proportional to the extent that feature's impact is global (

Load More