So after all these hours talking about AI, in these last five minutes I am going to talk about:

Horses.

Engines, steam engines, were invented in 1700.

And what followed was 200 years of steady improvement, with engines getting 20% better a decade.

For the first 120 years of that steady improvement, horses didn't notice at all.

Then, between 1930 and 1950, 90% of the horses in the US disappeared.

Progress in engines was steady. Equivalence to horses was sudden. But enough about horses. Let's talk about chess!

Folks started tracking computer chess in 1985.

And for the next 40 years, computer chess would improve by 50 Elo per year.

That meant in 2000, a human grandmaster could expect to win 90% of their games against a computer.

But ten years later, the same human grandmaster would lose 90% of their games against a computer.

Progress in chess was steady. Equivalence to humans was sudden.

🚨 I've a paper out today: Scaling Scaling Laws with Board Games! 🚨

arxiv.org/abs/2104.03113

Principle result is that by studying a sequence of small problems in ML, I could predict the outcome of experiments on orders-of-magnitude larger problems 🤯 I worked on Hex. Hex is a board game, with all the strategic depth of Go but also a much simpler rule set. Crucially, Hex on small boards is easy, and Hex on big boards is hard!

I can't recall any _techniques_ that knocked me off my chair, but there have been a couple of papers on training phenomena which have had a serious impact on how I think about RL:

https://twitter.com/jachiam0/status/1207975400929976320

'Meta learner's dynamics are unlike learners': you ask a regular NN to learn a transformation and it'll learn the component with the largest eigenvalue first, then the second largest, etc etc. A meta-learner will learn all the components simultaneously! arxiv.org/abs/1905.01320