iiinova — where your digital self keeps evolving

harry law

@harrylaw.bsky.social

1.4K followers 119 following 23 posts

Thinking about thinking machines | University of Cambridge and Leverhulme Centre for the Future of Intelligence | Previously Google DeepMind

Posts Replies Media

harry law @harrylaw.bsky.social · 30/05/2025

Every week I write a short history of AI: a trend, person, or technology from AI or the fields that influenced it This week deals with the Mechanical Turk. I look at its tricks, second life, and games with Ben Franklin, Edgar Allan Poe, and Napoleon www.learningfromexamples.com/p/the-man-be...

learningfromexamples.com

Please ignore the man behind the curtain

AI Histories #5: The Mechanical Turk

110

harry law @harrylaw.bsky.social · 18/03/2025

A bit on AI in the Victorian imagination by way of the AI researcher’s favourite classic book, Erewhon, and what came before it www.learningfromexamples.com/p/the-coming...

learningfromexamples.com

The Coming Race

Bovril Among the Machines

020

harry law @harrylaw.bsky.social · 04/03/2025

Definitional questions about AI can be tedious, but their staying power is curious. I argue that's because AI is the quintessentially Verhoevenian technology: an especially supple cypher for politics, values, beliefs, and biases Link below

120

harry law @harrylaw.bsky.social · 18/02/2025

I wrote about the territory of the moment. Link below for a history of science essay about Greenland, by way of pulp sci-fi outposts, lost colonies, and Sherlock Holmes 🧊

130

harry law @harrylaw.bsky.social · 04/02/2025

New blog on something I've been wanting to write about for a while: historical parallels between AI and alchemy, where they hold & where they break www.learningfromexamples.com/p/the-econom...

learningfromexamples.com

The Economy of Magic

On alchemy, imagination, and machine learning

140

harry law @harrylaw.bsky.social · 14/12/2024

extremely zoomable, super high resolution picture of de heem’s ‘flowers in a glass vase’ from 1660: tahpaul.github.io/deheem/

020

harry law @harrylaw.bsky.social · 30/11/2024

A Turing test for poetry, copyright, and finetuning - three things I enjoyed writing about in this week’s newsletter. Check it out if you’re into that sort of thing www.learningfromexamples.com/p/conversati...

learningfromexamples.com

Conversation Galante [TWIE]

The Week In Examples #61 | 30 November 2024

030

harry law @harrylaw.bsky.social · 29/11/2024

in 1971 you could look like this and play a 29 year old

040

harry law @harrylaw.bsky.social · 29/11/2024

guy who is sceptical of AI but supports export controls on GPUs

230

harry law @harrylaw.bsky.social · 26/11/2024

Colleagues at Google DeepMind wrote a new essay aiming to provide an overview of the AI for science opportunity, a nuanced take on the risks, insights on the ingredients needed to succeed, and an ambitious policy agenda. Read it below on our policy blog 🫡

170

harry law @harrylaw.bsky.social · 22/11/2024

it’s friday night

040

harry law @harrylaw.bsky.social · 18/11/2024

let me tell you about instrumental convergence, anon

160

harry law @harrylaw.bsky.social · 17/11/2024

Every week I publish a newsletter about AI and society. This week I looked at: 🛣️ Calls for ‘AI highways’ 🌍 Safety implications of world models ✍️The first draft of the EU’s Code of Practice for AI www.learningfromexamples.com/p/my-w…

learningfromexamples.com

My way or the highway [TWIE]

The Week In Examples #59 | 16 November 2024

130