Letter: How our own brains work may be AI’s biggest insight
![MRI Image Of Head Showing Brain](https://www-ft-com.ezproxy.brunel.ac.uk/__origami/service/image/v2/images/raw/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net%2Fproduction%2Feb9faaef-8e2c-475a-936a-61ee5e952d4f.jpg?source=next-article&fit=scale-down&quality=highest&width=700&dpr=1)
I agree with the motivation for your editorial (FT View, May 31), but think the artificial intelligence you speak of is a “black box”, whether it should be or not. It is trillions (at least — more likely several powers of 10) of artificial neurons, inspired by natural systems, interacting in ways we don’t fully understand. It is not a car with well-defined human components: engine, wheels, etc. Basel-type regulation is not applicable. We know the inputs: pretty much everything (eg the whole of Wikipedia).
We know how to build such systems but in some ways training them is akin to growing a new plant or creature. Whether they prove of commercial use or not, perhaps the most useful insight such models might provide is to know how our own brains work: biological systems are more difficult to study.
For the time being I shall ponder the wonderful mind of my wife. A trivial argument with her inspired this letter!
Roderick Wallace
Beckenham, Kent, UK
Comments