Man in the street sheds light on AI

Growing up with Lucy
October 8, 2004

Lucy is Steve Grand's attempt to understand mind and life by building an ape-like robot. He may not have created the robotic wonder of all time, but armed with a few "crummy electric motors" and a mechanised orang-utan mask, he has used the android as a vehicle on which to base his foray into the jungle of theories, philosophies and prejudices that surround the capture of human-like intelligence in an engineered machine.

Grand describes himself as a self-taught "man in the street" who works out what is needed from first principles rather than the pronouncements of the artificial intelligence gurus. This image is somewhat belied by his OBE, obtained for designing a most successful computer game, and his fellowship from the National Endowment for Science and Technology, which ensured that he could continue building Lucy in his shed while writing Growing up with Lucy .

The innocent-bystander stance will engage readers put off by the "how-clever-I-am" attitudes of some professionals. The approach also brims with educational value for the general reader. Grand explains a range of technical methods that help make a robot, how we communicate with our PC and how collections of little servo-motors can do interesting things. His explanations of mind are technological speculations on what the anatomy of the brain might be doing. The book glides easily over its major parts: body, mind and spirit. Much is devoted to the visual sense, particularly how the world might be represented in the visual cortex.

I realise that by criticising these speculations I risk Grand branding me as one of those unpalatable academics blinkered by professional ambitions.

But although Grand's intuitions work very well most of the time, one should beware of one or two minor digressions that contrast with what others have found. For example, he tries to explain how the early parts (V1 and V2) of the visual cortex might represent a solid, integrated visual world. From the work of Francis Crick and Christof Koch, current paradigms place this capacity in deeper brain regions (the extrastriate cortex) where the motor areas of the brain can help with compensation for eye, head and body movement.

But this is nit-picking. Grand's book is a delight, and carping academics should admire his ability to cover a vast amount of material on brain structure and good design practice. I was particularly taken, near the book's end, by some robust philosophy, which he has developed in the face of those who accused him of being an "unscientific charlatan", a "heartless reductionist" or of contributing to the destruction of humanity by marauding hoards of over-intelligent robots. Grand's view, which I share, is that in machines, stupidity is to be feared a great deal more than intelligence. He wrote some of this the day after 9/11. Clearly evil minds can harm humanity without building smart robots. Rusty rifles, hand grenades or normally benign aircraft will do.

So, Growing up with Lucy should be read to get a sane, level-headed view of intelligent artefacts. Grand's sharp mind proves that even a few "crummy electric motors" put together in the garden shed go a long way towards addressing age-old body-mind puzzles.

Igor Aleksander is emeritus professor of neural systems engineering, Imperial College London.

Growing up with Lucy: How to Build an Android in Twenty Easy Steps

Author - Steve Grand
Publisher - Weidenfeld and Nicolson
Pages - 229
Price - £16.99
ISBN - 0 297 60733 2

Please login or register to read this article.

Register to continue

Get a month's unlimited access to THE content online. Just register and complete your career summary.

Registration is free and only takes a moment. Once registered you can read a total of 3 articles each month, plus:

  • Sign up for the editor's highlights
  • Receive World University Rankings news first
  • Get job alerts, shortlist jobs and save job searches
  • Participate in reader discussions and post comments

Have your say

Log in or register to post comments

Most commented