Adrian Berry  
Science author and columnist   
Home | About Adrian | Subjects | Contact Me | Reviews | Links 
 

We're Better than Robots

IMAGINE a mining engineer who was given the following four tasks, and estimate whether he has any chance of being able to perform them:

(1) drill down through two miles of solid rock, and extract samples of whatever lies beneath.

(2) Prepare and execute the drilling operation entirely with robots, with no humans on the spot. If any difficulties arise, the robots must receive fresh instructions from humans located at a vast distance from the site.

(3) Make allowance for the fact that receipt of these instructions will always be delayed by at least four minutes, and sometimes by as much as 22 minutes.

(4) In the event of a single major error, be prepared to tell your bosses that you have spent hundreds of millions of dollars without achieving anything, and with no hope of recovering a cent of it.


Solar Systems could be explored by robots alone?


These are the tasks confronting scientists as they plan to drill beneath the crust of Mars in their search for bacterial life. The majority of them have finally admitted that it could not be done in the way described above.

The welcome news, in short, is that there is now a scientific consensus, shared by NASA, that astronauts will eventually be needed for the search for life on Mars - a quest upon which the space agency has set its heart. While robots can perform preliminary searches on the Martian surface, the really difficult parts of the project can only be carried out by humans.

This is particularly welcome news since, as recently as a year ago, many scientists were insisting that the entire Solar System could be explored by robots alone. It was always a vain prospect, but academics seemed unable to understand that machines are not clever enough - and may never be clever enough - to carry out tasks where unexpected and formidable obstacles are liable to arise. Solving them requires common sense and imagination, qualities that machines wholly lack.

The same will apply to plans to search for life in the ocean that is believed to exist beneath layers of ice on Jupiter's moon Europa. The proposal is that a submarine should be lowered to Europa's surface, that a hole should be punched in the ice, and that the craft be launched into the ocean to carry out the search. I cannot believe that it will be possible to conduct effectively such a vast operation on Europa, a world that seldom comes closer to Earth than 400 million miles, simply by pushing buttons in a control room in Pasadena.

The trouble with our computers-can-be-taught-to-do-anything outlook is that people tend to forget that the cyber world is much simpler than reality.

One could easily simulate a search for oceanic life in Europa in a computer game with convincing graphics, but the graphics would all be clever fakery irrelevant to the real Europa, the real distance from it, the real ice, the real ocean, the real submarine and the real problems.

 

Home | About Adrian | Subjects | Contact Me | Reviews | Links