Should you discount someone’s defense of AI because they get money from AI investors? Does a lack of financial stake make one more objective? Not necessarily.
joshdmay.substack.com/p/conflicts-...
Really thoughtful review of the Neuroethics book. Elzė Sigutė Mikalonytė & Brian D. Earp describe it as “nuanced, insightful and well-considered”—though they also raise some interesting concerns, well-captured by their review’s clever title. www.tandfonline.com/doi/full/10....
Some arguments against AI are solid. Others… not so much. In this adversarial collaboration, we attempt to sort the arguments into each box and recognize our own biases in the process.
open.substack.com/pub/joshdmay...
Not an April Fool’s post, unfortunately. www.nature.com/articles/d41...
“In the same way that you’re unlikely to eat Twinkies as a regular snack or still believe that Pop-Tarts provide a balanced breakfast, stop consuming ultraprocessed content.”
Provocative article by the great Cal Newport. Love the analogy to ultrprocessed foods.
Got interviewed for an episode of this great podcast, Unexplainable by Vox, which is about "everything we don't yet know." Perfect theme for an episode about neuroscience, free will, and the law.
open.spotify.com/episode/5N0Q...
New paper out in Nature with some folks at Google DeepMind. We argue that AI systems should be evaluated for their *moral reasoning* just as they are evaluated for their ability to code or do math. And we provide a roadmap for evaluating moral competence in LLMs.
www.nature.com/articles/s41...
Is a targeted, month-long national economic strike an effective form of protest against the unchecked power of masked agents who treat the Bill of Rights as a suggestion? I don’t know, but it sure is an opportunity to rethink where my money is going.
I get asked a lot about how universities are dealing with AI. We can cope, I think. But we need to tread carefully and get back to basics.
open.substack.com/pub/joshdmay...
AI boosters and critics should focus more on the arguments
