Robot Scientist meets the press

I am writing this at lunchtime on Thursday but can't post 'til midnight because of the embargo (ah Clive, how well you trained me!). The Robot Scientist paper comes out on the Science website tomorrow, and in the journal soon after.

My former colleagues at Aberystwyth have been busy giving interviews and being filmed/photographed for the last couple of days. There's a science wire service I'm still subscribed to after the BSA Media Fellowship and I'm really excited, because they've just put out the first comment I've seen from someone not on the team!

I can't wait to see the coverage tomorrow. I have no idea how well the story will be treated - but I did suggest a headline for the press briefing paper: "Robot Scientist beats humans to new knowledge" so I hope that angle gets covered, rather than robots to put scientists on dole/cure cancer/kill everyone with GM yeast.

Using AI, knowledge about the domain and a lot of lab automation kit the Robot Scientist managed to discover (for itself) the previously unknown function of 12 genes in brewer's yeast. The robot starts with some knowledge about the yeast metabolism and is allowed to design experiments on yeast with genes knocked out. It uses what it learns from those experiments to design another set of experiments, then another, until it finally figures out what the gene does.

With more than 6,000 interesting genes in yeast alone, that would take humans years: the robot can run 24/7 (in theory) and run thousands of experiments in parallel. If the robot just did those experiments combinatorially, that would still take more years than we're likely to have left on this planet - so the AI for hypothesis generation and experiment planning really comes in handy!

The paper.

UPDATE: My favourite writeup is this from Clive Cookson at the FT. It's sober, well informed and distills what the work really achieved. I'm defnitely buying a copy of the pink 'un today!

One of the best writeups so far, from the land down under: Sydney Morning Herald

MSNBC also has a pretty good take on it.

Auntie Beeb has done a good job too.

And my personal favourite writeup from a blogger.

It thinks with its brain now

Jeff Hawkins is one of those jaw dropping people who seem to have an amazing capacity to make things happen. He's the pioneer behind "pen based computing" and the Palm and Handspring companies.

More recently he's used his money and inventiveness to found the Redwood Centre for Theoretical Neuroscience, based at the Helen Wills Neuroscience Institute at Berkeley. From this, along with long term collaborator Donna Dubinsky, he's spun off a company called Numenta.

Why the business update? Well, Hawkins is of the view that current research in AI will never succeed in emulating "holistic" human intelligence. In that he's right. In the AI community, we're very good at solving particular problems (such as how to win at chess) and developing particular techniques (like the neural net or genetic algorithm). We have computers that are a lot smarter than computers "ought" to be - this area of research is known as "weak AI". We're also pretty good at using some of these intelligent modes of computing to better understand living intelligences, like human perception for instance.

Strong AI - making a computer that thinks like a human - has been a dirty word for some time now. The reins on this area of research were tugged in pretty sharply at the onset of the AI winter and have never been slackened since.

Hawkins may be about to change all that with his model of Hierarchical Temporal Memory (HTM). The details of his group's work can be found here and [PDF] here.

Hence the business update. I get my jollies from evolutionary computation and logic programming, so this is not my area. From the (very) little I do know about neurally inspired computing, this doesn't look like "neural nets 2.0". But I don't know: is this old news, a pipe dream, or the shot in the arm that brings "hard AI" back to life?

If I read something like this from any other entrepreneur I'd be cynical as all getout. On Hawkins' past form however, I'm downgrading that to merely sceptical. With an option on secretly hoping to be convinced.

And if all that was just too dry for a Friday - take a look at this "tutorial" on neural architecture:



(Spotted on Mind Hacks