Thursday, November 29, 2007

Hugh Everett III and J.P.Morgan

The December issue of Scientific American contains a feature on the life of Hugh Everett III, the inventor of the many-worlds interpretation of quantum theory. Everett left academia in a state of some bitterness at the initial lack of enthusiasm for his proposal. After a period of time working for the Pentagon on the mathematics of nuclear warfare, he established Lambda, a private defence research company. Under contract to the Pentagon, Lambda used Bayesian methods to develop a system for tracking ballistic missiles. Most intriguingly, though, the article suggests that Everett may have subsequently hoodwinked American bank J.P.Morgan:

John Y. Barry, a former colleague of Everett's...questioned his ethics. In the mid-1970s Barry convinced his employers at J.P. Morgan to hire Everett to develop a Bayesian method of predicting movement in the stock market. By several accounts, Everett succeeded - and then refused to turn the product over to J.P.Morgan. "He used us," Barry recalls. "[He was] a brilliant, innovative, slippery, untrustworthy, probably alcoholic individual."

This information appears to have been taken from an on-line biography of Everett, which quotes Barry as follows:

"In the middle 1970s I was in the basic research group of J. P. Morgan and hired Lambda Corporation to develop...the Bayesian stock market timer. He refused to give us the computer code and insisted that Lambda be paid for market forecasts. Morgan could have sued Lambda for the code under the legal precedent of 'work for hire'. Rather than do so, we decided to have nothing more to do with Lambda because they, Hugh, were so unethical. We found that he later used the work developed with Morgan money as a basis for systems sold to the Federal Government. He used us...In brief a brilliant, innovative, slippery, untrustworthy, probably alcoholic, individual."

What I don't understand here, is how this recollection is supposed to substantiate the claim that Everett had dubious ethics. Within an economy, there are some companies which provide the fundamental level of production within the economy, companies which invent things, design things, develop things, discover things, and make things; and then there are parasitic companies, such as banks and firms of lawyers, which produce nothing, and merely drain their wealth from the activities of others. Taking money from a large American bank to subsidise a research project, and then preventing that bank from harvesting the fruits of that project, is the very paragon of ethical behaviour.


Dave said...

Interesting blog entry Gordon. It seems that because Everett was
employed under Lambda Corporation, he was able to keep his code rather than handing it over to J P Morgan. Odd that they didn't sue him - they must have felt that it wasn't worth the bother, for whatever reason.
If he had been directly employed by J P Morgan I doubt he would have been able to do this. These days many large companies have in their contract of employment a clause stating that anything created whilst being employed by them is owned by them.
Not sure how enforceable this in in practice however it is obviously
there for a reason.
Outrageous really.

Gordon McCabe said...

Cheers Dave.

I can't help thinking that if Everett had been clearly in breach of contract, then JP Morgan would indeed have sought restitution in the courts. In Barry's testimony, it's perhaps significant that Everett requested payment for the market forecasts produced using his software. Were JP Morgan buying just the forecasts, or the software as well?

Dave said...

Sounds to me like they were buying just the forecasts and not the software. As you say they would have gone to court - banks have very deep pockets and would usually settle out of court or not pursue the case if it was at all unclear - e.g. in J P Morgan's case $2bn - Enron, $2bn Worldcom etc.