[alicebot-aiethics] Computers vs. stock brokers/analysts ...

Richard Wallace alicebot-aiethics@list.alicebot.org
Fri, 24 Aug 2001 20:20:17 -0700


> And it's only a matter of time before we have a bot for a CEO running a
company,
> then a Congressman, then the President, then well who knows ;-)
> I think IBM even recently posted that a bot could make smarter, better,
and more
> profitable decisions than a stock broker/analysts

I used to work on vision systems for self-driving cars and trucks (and yes,
tanks) back in the 1980's.  We realized then that automated car does not
have to be perfect to be practical. It can still have accidents.  It only
has to be slightly better than human drivers, not perfect.  As long as robot
cars crash less frequently than human driven ones, the insurance companies
will insist on robot cars, or at least higher premiums for human drivers.

The point at which automation becomes successful is roughly the same for all
human jobs.  Once the robot becomes only marginally better than the human
decision maker, the robot will be the only practical, economic choice.
The robot stockbroker can still lose money, even a lot of money, yet perform
better on average, than the human broker.

It's never too early to start a presidential campaign.  ALICE for President
2004!

Rich