RYviewpoint

I always get a laugh at the famous Groucho Marx line "Who do you believe? Me or your own lying eyes?"

Well, here is a claim reported by the BBC of a major advance in AI (Artificial Intelligence) with the creation of a "robot scientist".

Don't believe me? See for yourself, with your own lying eyes:

From the abstract to the research report published by a team at Aberystwyth University:

We report the development of the Robot Scientist “Adam” which advances the automation of both. Adam has autonomously generated functional genomics hypotheses about the yeast Saccharomyces cerevisiae, and experimentally confirmed these hypotheses using laboratory automation. We have manually confirmed Adam's conclusions using additional experiments. To describe Adam's experiments we have developed an ontology and logical language. The resulting formalisation involves over 10,000 different research units in a nested treelike structure, ten levels deep, that relates the 6.6 million biomass measurements to their logical description. This formalisation describes novel scientific knowledge discovered by a machine.

Sounds impressive.

However, if you read to the bottom of the paper, you discover the typical qualifications in claims about major advances in "intelligent" systems:

Adam is a prototype and could be greatly improved. Adam's hardware and software are “brittle”, so although Adam is capable of running for a few days without human intervention, it is advisable to have a technician nearby in case of problems. The integration of Adam's AI software also needs to be enhanced so that it works seamlessly. To extend Adam we have developed software to enable external users to propose hypotheses and experiments, and we plan to automatically publish the logical descriptions of automated experiments.

So... this giant has feet of clay. This "robot scientist" is a lot like the Wizard of Oz. Just off to the side is a technician madly pulling levers trying to keep this "brittle" bit of software from falling apart and there is a team of scientists hovering nearby providing a little bit of scientific "support" in this machines scientific hypothesizing.

While I'm throwing cold water on the team's claims, I'm not dismissing the work. This is yet another small step in building more impressive tools with greater intelligence and autonomy. But sadly it is another example of AI researchers inflating their claims to get attention. This system will probably prove to be a useful aid in the laboratory as it is improved over time. It certainly shows that machines are getting more adept in many areas. But don't hold your breath that this machine, Adam, will step out of the lab and take charge of business or even figure out how to pay your taxes. This is a tool of limited scope with some intelligence. It would have been marvelous to demonstrate in 1960, but it falls short of being an honest example of a "robot scientist" in 2009.

As for the opinion of a working scientist, somebody doing drug research, here is Derek Lowe's viewpoint expressed in a blog entry:

But I remember the first time I saw an automated synthetic apparatus, back at an ACS meeting in the mid-1980s. There was a video in the presentation (a real rarity back then), and it showed this Zymark arm being run to set up an array of reactions, assay each of them after an overnight run, and report on the one that performed the best. “Holy cow”, I thought, “someone’s invented the mechanical grad student”. Being a grad student at the time, I wasn’t so sure what I thought about that.

...

I'm not planning on being replaced any time soon. But the folks cranking out the parallel libraries, the methyl-ethyl-butyl-futile stuff, they might need to look over their shoulders a bit sooner. That's outsourcing if you like - from the US to China and India, and from there to the robots.