On Cooperation

AI is very good at retrieving knowledge and reasoning about formal problems, but many of the issues that we face today require a different sort of skill—the ability to bring people together to cooperate. We have yet to see much evidence that AI is good at this—although there have been some early attempts, including one that I led at Deepmind. [A study on whether AI can function as a political mediator.] The ability to solve formal problems, like playing Go at Grandmaster level or solving logic puzzles, does not necessarily equip you for the real world, where things tend to be quite messy. Outcomes are uncertain—stochastic rather than deterministic—and challenges can arise without warning.

Nick Hildon - https://nautil.us/ai-has-already-run-us-over-the-cliff-1217080/. This is Chris Summerfield’s response to the question “Many people envision a near future where AI solves the looming issues of the day, but as you discuss in the book, this expectation may be misplaced. Why?” This is an interesting summary of Summerfield’s book, These Strange New Minds.

timothywstanley@me.com

I am a Senior Lecturer in the School of Humanities, Creative Industries and Social Sciences at the University of Newcastle, Australia, where I teach and research topics in philosophy, religion and ethics.

www.timothywstanley.com
Previous
Previous

On Humboldt’s World

Next
Next

On Kierkegaard’s Birthday