In 10 Years, Nobody Will Be Able to Lie (Part 1)

Within 10 years, nobody will be able to lie. Or more precisely, all lies will be detected instantly, so people will stop lying. Can you imagine how that will change society!?

But hold on a second! How is this going to happen?

Think about it: the technology is all here. We’re only waiting for someone to put the following pieces together.

Speech Recognition

My Samsung Galaxy Note 5 smart phone, now two generations old, does an astonishing job of recognizing speech. I can speak anything into it, on any subject, using my normal voice, and it is able to turn my speech into text nearly as reliably as a human could.

(And incidentally, Google recently announced big improvements in its language-translation software. It is now nearly as good as professional human translators.)

Semantic Understanding

Five years ago, IBM’s Watson computer, having digested 200,000,000 pages of content, including all of Wikipedia, trounced the all-time best human players of Jeopardy!.

Watson is optimized to answer questions. If my smartphone could somehow access a Watson-powered computer in the cloud, then I could speak any question into my phone, let my phone turn speech into text, and get an answer back from Watson.

But we don’t have to wait for Watson to be generally available. I have played out the above scenario countless times in actual conversations using nothing more than Google. For example, last night I was with some friends and we were wondering whether chimpanzees are of the same genus as humans. I whipped out my phone and asked it, “What are the family, genus and species of chimpanzees?” Sammi (my name for the woman inside my phone) understood my question perfectly and up popped a Wikipedia article with the answer. (They are of the same family, Hominidae, but of genus Pan rather than Homo.)

I kid you not: I have gotten to the point that I don’t enter any social situation without my phone, not because I intend to talk on it but because it is so often useful to get a quick answer to any question that may come up.

Semantic Comparison

Google has developed a tool with the whimsical name Parsey McParseface that can detect inconsistencies of meaning within a body of text. Such a tool could also be brought to bear in our task of real-time lie detection.

Suppose someone has just said to us, “Chimpanzees are humans, too!” Our lie detector could turn her speech into text, summon articles from known-reliable sources that define what it means to be human and what it means to be a chimpanzee. It would combine them with what our interlocutor said and ask Parsey McParseface if there are any inconsistencies. It could whisper the result through a wireless earpiece that we would be in the habit of wearing. Using today’s technology, this could all happen before our conversation partner had finished her next sentence.

Visual and Aural Clues

But it gets better. Earlier this year, it was reported that Chinese scientists have developed a lie detector based on facial and aural clues. Attach both an inconspicuous camera and a concealed microphone to our hypothetical device and it will be able to analyze tone of voice and facial patterns of whoever we’re talking to, as well as the semantics of his words. How powerful will that be!?

Affordability

As I said, all of the pieces are available now, although not all are in commercial production. But products of this sort come to market very quickly. There’s no need for long government approval processes (as with pharmaceuticals), and very little consumption of raw materials. It’s mostly software.

But affordability is the real story. Technology that we would have had to pay for 10 years ago is now free.  You used to pay for speech-recognition software that was of only marginal utility; now excellent speech recognition comes on our phones for free. You used to have to spend hundreds of dollars for a multi-volume encyclopedia; now you have Wikipedia, which is just as accurate but free. (If you can, keep it that way for everyone by donating to it!) You used to pay big money for a daily subscription to a newspaper; now major news sites with excellent reputations can be accessed for free.

Ray Kurzweil, of Singularity fame and now a Director of Engineering at Google, has observed that technology starts out being expensive and barely usable, and then progresses to being excellent and nearly free. This will be as true of lie-detection technology as it has been for long-distance telephony, computer memory, flat-screen computer monitors, and the Internet itself.

The ability to detect the lies all around us is surely as important to most people as the ability to talk on the phone. I think lie detectors will be as ubiquitous in 10 years as smartphones are now.

The Social Implications

So how do you think society will change when we can all tell whether each other is lying? That will be the subject of the next post.