White Matter

Science Today

Web edition, published July 21

What were they thinking?

Doubts cast on mind scanner claims

If it sounds like science fiction, quite possibly it is.

In 1989 Pons and Fleischmann secured a place in infamy by suggesting they had achieved nuclear fusion at room temperature. In 2004 South Korean researcher Hwang Woo-suk announced the successful cloning of human cells. What these two claims had in common was that both were wrong. Dr. Hwang's breakthrough, it turned out, was bare-faced fraud involving the falsification of data. History's verdict on Pons and Fleischmann has been more generous, seeing their crime as not motivated by an intent to deceive, rather an overly optimistic interpretation of ambiguous experimental results.

To this ignominious pair can we now add a third? At a press conference last week Lance Coriolis and his team at an obscure technology firm with the ominous name of Spurious Developments astounded the world by demonstrating a scanner they claimed was capable of decoding human thought. A mind reading machine, in other words.

Too good to be true? In a statement released today, experts granted access to the machine have cast serious doubts over its ability to replicate its maker's claims.

"It just doesn't work," said lead reviewer Hermann Kohn. "It lacks the resolution needed to pick out individual axon connections, while its algorithms for interpreting those features it can detect would best be described as crude. They are certainly inadequate for producing the realtime trace of human thought that the developers claim is possible."

This point about resolution is crucial. Without a sufficiently crisp image of the circuits of a brain, trying to decode what it is thinking is like trying to read a cheaply printed paperback novel that has been left out in the rain, the ink all run together in an illegible blot. If the machine is incapable even of reading the letters, there is no way it can be expected to make out the words.

So is this a case of wilful misrepresentation, like Hwang's clone, or simply the premature enthusiasm of cold fusion? The answer appears to be a mixture of the two. The expert reviewers are unanimous in their assessment that the machine does indeed represent a major step forward in neural scanning technology. But they also agree that, impressive though its performance may be, nobody with any understanding of the subject would be fooled for long about its ability to solve the much more complex problem of mind reading.

This leaves open the question of why Spurious Developments would make such a claim in the first place, and then, having done so, allow experts to examine the machine in the near certainty that the hoax would be uncovered.

Financial fraud is one possible explanation, but Spurious Developments is not seeking new investment and claims to have fully pre-funded its development program for some time to come. Nor is it an attempt to manipulate the share price. In fact, the company's stock is closely held by a consortium of long-term venture financiers, and no actual share sales have taken place in the periods leading up to and after the announcement. With no transactions, there has been no price to manipulate.

Similarly, the company has not offered its technology for sale, and therefore has not violated any commercial laws on the making of misleading claims.

So the mystery remains: whatever could they have been thinking? For now all we can say with any certainty is that their brain scanner will not be much help in answering that particular question.