In 1997, IBM defeated Grand Master Gary Kasparov, the world chess champion, with its computer Deep Blue. It was one giant leap for computerkind in the development of artificial intelligence. The work continues at places like MIT, Cal Tech and IBM in the attempt to create a machine that can think the way we do.
Another attempt to prove--or disprove--the power of computers to think was the “Turing Test” developed by the British inventor Hugh Loebner.1 The test is named for the British mathematician Alan Turing, who, in 1950, was one of the first to address the question, Can a machine be made to think? In the Turing Test, a judge enters into a dialogue with a human or computer contestant at a remote location. After five minutes, the judge must decide which brain he has been talking to--the machine or the human. If the machine fools the judge, it earns the title of Most Human Computer, and the programmers win $3,000. If the human contestant wins, he earns the title of Most Human Human.
IBM’s latest attempt has been with its new system called Watson, which runs on about 2,500 parallel processors and can perform 33 billion operations every second.2 By feeding vast amounts of data and arcane information into the system--dictionaries, databases, histories, encyclopedias, the Bible and so on--the programmers have given Watson, presumably, all the knowledge a normal human being has stored in her brain, and more. But thinking requires more than mere data, so the creators of Watson have developed algorithms to enable something close to the thinking process in which the human mind sorts, compares, evaluates and selects the right data in order to make the right decision. To prove the strength of Watson’s A.I., IBM challenged the producers of the TV game show Jeopardy! to a series of matches in which some of the most successful participants of the past would compete against Watson.
As in the Turin Test, Watson must deal with syntax, puns, innuendoes, analogies, metaphors and whatever other tricks of speech recognition and communication may be thrown at it while competing in the Jeopardy! game. Logical analysis and analytical thinking are required.
If a computer program can be taught to operate with such cognitive ability, what is the limit to how such (artificial) intelligence may be applied?
Surveying is more than a mere exercise in measurement. We are concerned with evidence and analysis, systems of ownership and tenure, land management and cadastral systems, applications of “green” development and the economics of land use. From my own observations, seventy percent of a surveyor’s projects involve straightforward solutions. Twenty five percent require more deliberation, while the remaining five percent are so demanding as to require solutions that may end up in litigation and reversal of the surveyor’s decisions.
The eminent economist Paul Krugman observed in a recent New York Times editorial that “Computers, it turns out, can quickly analyze millions of documents, cheaply performing a task that used to require armies of lawyers and paralegals.”3 If that is so, could not such a system be fed every statute relating to land tenure, every court decision settling common boundary and ownership issues, all the rules of evidence, every set of technical and procedural rules for the practice of surveying and every expert commentary on surveying as well as all the geometry, trigonometry, least squares and network balancing routines ordinarily applied by surveyors? With this accumulation of knowledge, could the system not perform the same logical analysis the surveyor uses every time she reconciles evidence with the record in a solution of a point, corner, line or parcel definition? When/if that day comes, the surveyor’s task might, after all, be reduced to mere measurement. If the ancient and venerable legal profession can see its activities so reduced and simplified by the A.I. of the computer revolution, can the surveying profession hope to avoid the same fate? Will we no longer be challenged by that treacherous 5 percent of our projects?
Krugman made an additional observation: “It is a truth universally acknowledged that education is the key to economic success. Everyone knows that the jobs of the future will require ever higher levels of skill.” There is a flip side to all this, however, and it has to do with occupational opportunity. In his column, Krugman adds, “It’s no longer true that having a college degree guarantees that you’ll get a good job, and it’s becoming less true with every passing decade.” He adds that giving workers college degrees “may be no more than tickets to jobs that don’t exist or don’t pay middle-class wages.”
In a distillation of all of the above, not withstanding our natural tendency to reject the unthinkable, it seems clear that our profession will continue to change, sometimes with disturbing results. To stay ahead of the technological tsunami, our educators must be futurists, our practitioners must be realists, and our entry-level surveyors must be educated to the max, ready to adjust their career goals--and they had all better be quick on their feet, prepared to stay one step ahead of the changes.
References1. Christian, Brian, “Mind vs. Machine,” The Atlantic Magazine, March 2011, www.theatlantic.com/magazine/archive/2011/03/mind-vs-machine/8386/.
2. See www-03.ibm.com/innovation/us/watson/ and www.nytimes.com/2011/02/ 17/science/17jeopardy-watson.html. For the results of 55 matches on “Jeopardy! The IBM Challenge,” Google Jeopardy Watson.
Krugman, Paul, “Degrees and Dollars,” International Herald Tribune (NY Times), March 6, 2011, www.nytimes.com/2011/03/07/opinion/07krugman.html.