Ben Cravens

International treaties on AI should only target inherently dangerous models, not tools

The crack in the red lines, Kantian AI regulations, and why pragmatism beats purity

Nov 7, 2025 • Ben Cravens

August 2025

On Scaling's End

The end of pretraining scaling; chart crime; post-training boosts math and coding; economic risks; bubble talk

Aug 29, 2025 • Ben Cravens

On Not Predicting the Future

Why discussions about AI's future often miss the point

Aug 3, 2025 • Ben Cravens

June 2025

New Zealand Should Establish an AI Safety Institute

We are currently lagging behind other developed nations on this critical issue.

Jun 1, 2025 • Ben Cravens

1

March 2025

Debunking Arguments against the Australian Social Media Ban

Social media sucks actually, and it is good that Australia is banning it for teenagers

Mar 27, 2025 • Ben Cravens

AI in the State of Nature

Why global problems like AI require global governance

Mar 22, 2025 • Ben Cravens

November 2024

Simple Steps to Avoid the Attention Economy

Modern problems require modern solutions

Nov 15, 2024 • Ben Cravens

1

May 2024

Ok, Doomer

How AI is making us all Doomers

May 3, 2024 • Ben Cravens

November 2023

New Zealand: The incoming ban on phone use in schools doesn't go far enough

Pairing it with a ban on adolescent use of social media would be more effective

Nov 9, 2023 • Ben Cravens

1

July 2023

Is openness in AI development a good thing?

It's complicated and depends on your faith in institutions

Jul 28, 2023 • Ben Cravens

2

February 2023

The future of this substack (meta)

Some housekeeping

Feb 26, 2023 • Ben Cravens

Large Language Models explained: part 1

What is a transformer? (the "T" in Chat-GPT)

Feb 18, 2023 • Ben Cravens