Vice Provost & Dean @NUSingapore/@NUSCollege; @NUS_Law; Director @AISingapore | AI & data | international law | occasional fiction | the odd New Yorker cartoon
Singapore
Joined August 2011
- Tweets6.6K
- Following4.6K
- Followers5.5K
- Likes5.8K
Pinned Tweet
I’m thrilled that my first work of general fiction, “Artifice”, is at last available worldwide! Skip the thread and order the book here: simonchesterman.com/2023/06/…
Or read on! #Artifice
Call my agents! The news around AI has been even more frenetic than usual. On the one hand, agentic AI is misaligned and coming to kill us all. On the other, AI is solving maths problems not just quickly but elegantly.
The basics should be simple: No accountable human; no deployment. No meaningful possibility of containment: no development at all.
Plus, I managed to get in a reference to the Butlerian Jihad for all the Dune nerds out there. Enjoy! straitstimes.com/opinion/whe…
The Straits Times is paywalled — and you should pay for journalism, right? (Right?) — but the piece is also accessible on my personal site here: simonchesterman.com/2026/10/…
#AI #AIGovernance
One of the ideas at the heart of NUS College is that education should sometimes involve breaking open the classroom — learning not only from books and teachers, but from communities, experience, failure, and one another.
Perhaps my favourite line in the story comes from one of our students, describing the goal of their project as making themselves “obsolete”. That is a pretty good test of genuine impact: not what happens while you are there, but what continues after you leave.
Very proud of our students, colleagues and — especially — the community partners who make this kind of learning possible.
#NUSCollege #NUSC #ImpactExperience #ExperientialLearning news.nus.edu.sg/how-nus-coll…
Now that AI safety has moved from academic articles and the after-work drinks of computer engineers to the front pages of newspapers and the delegates’ lounge at the UN, the question of what global institutions might actually help govern AI is becoming more concrete.
But many of the most consequential decisions about frontier AI remain concentrated in a very small number of private companies.
Perhaps the emerging architecture of global AI governance is not architecture without an architect, so much as architecture designed by multiple actors with different forms of power, divergent incentives, and no common set of plans. simonchesterman.com/2026/10/…
“It’s almost like a constitution,” President Trump says. Yes: a voluntary, one-page document in which the world’s largest AI companies promise to self-regulate. Just like a constitution.
(Although constitutions generally spell the name of the country correctly.)
The days of letting our new silicon sovereigns govern themselves are over. straitstimes.com/world/unite…
“Never waste a good crisis” is often invoked when confronting bureaucratic barriers to change. If we are very lucky, recent crises in AI will prove serious enough to stir us into action. Otherwise, we may have to wait for something worse.
And for more of my own take on this: doi.org/10.1017/ajil.2025.10…
