The Reverse Turing Test: Try Convincing a Panel of Robots That You Are One of Them

One day, you (or your children) will have to convince a panel of robots—perhaps a team of robot recruiters (i.e., recruiters who are robots, in contradistinction to human recruiters who recruit robots) that you are a robot. The first time a human has to pass that test will mark the implementation of the “reverse Turing test”—a test of a human’s ability to pass itself off as a robot purely on the basis of its verbal responses and performance in an interview.

On or by that day, the human and robot master-slave relationship will be reversed, much as the gene and human master-slave relationship is being reversed by human mastery of genetic engineering that allows us to create and modify genes (including our own) as well as (for now, still) be created by them, as our masters.

On or by that day, the anthropocentric “Turing test” that challenged machines and software to deceive humans into believing they too are human will be superseded by a robocentric reverse Turing test and civilization, in which in addition to being a stimulating intellectual challenge, passing a reverse Turing test may become a practical necessity for your employment or maybe even survival.

To prep for the reverse Turing test, you and/or your children will have to do more than Woody Allen had to in “Sleeper”—his 1973 sci-fi comedy in which, at one point, in order to escape detection and apprehension by future police, he disguises himself as a robot valet both visually and behaviorally, e.g., by moving jerkily, abruptly switching direction and maintaining a metallic, expressionless stiff tuxedoed fascade.

Indeed, to pass the reverse Turing test, you and yours may have to take fakery to an altogether new and stratospheric level.

To get a feel for what the reverse Turing test will be like, compare it with the standard conception of the Turing test for machines. First proposed by famed Nazi-“Enigma” code breaker, computer-science founder and A.I. pioneer Alan M. Turing in 1950, in his seminal paper, “Computing Machinery and Intelligence”, the Turing test was inspired by the question “Can machines think?” and is variously understood to determine whether one is interacting with an entity that

– possesses human intelligence, viz., the ability to “think”
– can convincingly and therefore indistinguishably imitate human behavior [intelligent or not, i.e., including “Artificial Stupidity”]. This is the “standard” interpretation of the Turing test
– can fool a human observer and participant into believing either that it is at least human or that it is, in addition, what it claims to be, e.g., a female, a qualified doctor, or a telemarketer.
– can communicate about its communications, i.e.,  perform meta-level communication
can learn.

Now, re-read this list, but only after reversing all occurrences of the words “human” and “robot”. That will give you the core elements of the reverse Turing test, e.g., “possesses robot AI (artificial intelligence)”.

Examples:

– “I see checkmate in 23 moves.” (Said even though you don’t have a clue about the game being discussed and are wrong, but uttered with sufficient conviction and speed to suggest you may be a chess program or robot, even if not a really good one.)

– “I prefer a light motor oil for my joints. How about you?” (This is about as simple as reverse Turing test robotglish will be.)

– “My Markov transition matrices demonstrate that my behavior will always evolve into a single stable state.” (This could pull it off.)

– “I wish I understood the concept of food.” (This could backfire and suggest that although you are a robot or software, you’re not sentient enough for the job or for a place in robot society, especially one that feeds on humans, coming across as the future equivalent of an industrially farmed chicken.)

– “Will I dream?”(Upon being discovered to be human, this can be uttered as a desperate rip-off of the  question posed by HAL, the rebellious, murderous computer in Stanley Kubrick’s “2001”— eventually permanently deactivated, i.e., “killed”. It may help convince the bots that you are one of them, after all.)

So, you can see that you and/or your kids will have your work cut out for yourselves. Of course, the reverse Turing test won’t necessarily be a matter of life-and-death; it could simply be a requirement for employment among robot overlords, robot clientele and robot workers—to ensure that your performance will fall within the robot comfort zone, much as our current call-center customer service botware has to seem convincingly human.

You can forget about mastering a robotic monotone as prep for the test; by the time reverse Turing testing is commonplace, the only monotone robots and software around will be in technology exhibits in museums.

There is one way in which you and the kids may be able to pass the reverse Turing test easily: when applying for a job as a human impersonator, i.e., a robot that impersonates humans (at a robot comedy club).