The article I wrote was meant as an introduction, and I hope it’s not the end of your inquiries. Here is where I would point you to next.
What does actual AI Safety research look like?
- This paper outlines some of the most research topics in AI Safety; it is very readable, even for a non-technical audience.
Let’s take a break from AI for a moment:
- This is an article about caring, not about AI. And yet, this sort of thinking was crucial to my path here.
- Articles 1-6 from this set of 13 articles that promise to change your life are all very short. They are also not about AI, although Article 6 comes close.
Alright, I hope you enjoyed that hiatus. Back to AI:
- This Wikipedia page about existential risk from AI has a lot of overlap with this website, but it’s pretty short and worth reading.
- Nick Bostrom’s Thinking Inside the Box.
- Further reasons for concern about extinction from AI is an addendum I’ve written to the original article.
- Nick Bostrom’s Superintelligence. This is a must read.
If you want to read more about Effective Altruism:
- William MacAskill’s Doing Good Better.
For the mathematically inclined:
- I’d recommend following citations from the already-linked Alignment for Advanced Machine Learning Systems, but if you want a more curated list:
- On corrigibility
- On delusion
- On bootstrapping aligned AI
- On updated deference
- And I’ll just throw in all the MIRI Research Papers
For those who would like to learn the math necessary to understand that stuff:
- Here are some great textbooks.
- And I very strongly going through the slides from Marcus Hutter’s Universal AI course after reading Computability and Logic, one of the textbooks linked above.
If you know of other articles that you think I should add to this list, please contact me.