Maybe, three? :-))) Oh, what a fruitful conversation -).Ĭlearly, Turing's original test has not withstood the test of time. Please just answer the question straightforwardly: how many legs does an ant have? SCOTT: No, I need to know that you're not a chatbot. > EUGENE: Just two, but Chernobyl mutants may have them up to five. SCOTT: How many legs does a millipede have? Maybe, three? :-))) By the way, I still don't know your specialty-or, possibly, I've missed it? And I forgot to ask you where you are from. > EUGENE: I can't make a choice right now. SCOTT: Which is bigger, a shoebox or Mount Everest? Consider this dialogue, which computer scientist Scott Aaronson had with the bot: In reality, Eugene knew-and understood-almost nothing. But so what? The judges got only a few minutes to conduct their interviews, and they undoubtedly would have ferreted out the truth soon enough. Some judges were fooled (about a third, enough to count as a shallow victory). In 2014, for example, a chat bot named “Eugene Goostman” supposedly “won” the Turing test by pretending to be a sarcastic 13-year-old boy from Odessa, Ukraine, who never answered questions directly. The way to win the Turing test is to deceive. But that odd twist is not an exception-it is the rule. If a judge asks a machine, “Do you have feelings?” the machine must lie to win. As Turing himself partly anticipated, the trick to winning is mostly not to answer the questions. One can “win” simply by being deceptive or feigning ignorance.
The problem is that Turing's test is too easily gamed. It is now possible to build machines that fool people, at least for brief periods-but the victories are fleeting, and they do not seem to be carrying us even close to genuine intelligence.
But it shouldn't be: the Rubicon can be passed, though for the wrong reasons.
Today, in the eyes of the general public, the Turing test is often seen as a kind of a Rubicon, a measure of whether machines have truly arrived. Anticipating what we now call chat bots-computer programs that masquerade as humans-Turing envisioned a contest in which a machine tries to trick an interrogator into believing it is human, answering questions about poetry and deliberately making mistakes about arithmetic. He called it the “imitation game,” but most people know it as the Turing test. In 1950 Alan Turing devised a thought experiment that has since been revered as the ultimate test of machine intelligence.
0 kommentar(er)
