How Gamers Might Demonstrate Why AI is Impossible (And Your Marketing Job is Safe).

We've been warned about it: the white-collar apocalypse is on the way. Intellectual labour is under threat. Jobs in fields like medicine, banking, journalism and marketing - yes marketing - are about to disappear in a puff of digital smoke. While outsourcing and the assembly line decimated (western) skilled labour in the 80s and 90s, the threat to (world-wide) intellectual labour will be very much in-sourced under our own noses: Artificial Intelligence (AI). We're staring down the barrel of being replaced by thinking machines. It's not a matter of if, the doom mongers say, but when. It's bound to happen.

But... is it?

It seems, if you read the science media, the answer is inevitably yes. But it's not going to happen. Why? Because AI is about replicating the mind and there's more to thinking than just the mind. Can machines really be capable of thinking? I don't mean processing, I mean thinking. Because that's what our work entails, that's what makes it "intellectual" labour, we don't just parse or process information, our jobs depend on empathy, understanding, wisdom and the judgement that that entails. Our business is language, the way we interact. Are machines capable of thinking like people? Could they speak to people?

No. What will forever confound machines is our use of language, between us, human to human.

Here's an example: L337 (“LEET”, short for “Elite”) Speak, has its origins in the bulletin board systems of the 1980s but really took off with the advent of search engines as hackers, crackers and gamers, the “elite” of the internet sought to disguise their messages from search engine bots and bulletin board filters. According to  Bruce Sterling, a historian of hacking, L337 can be traced back to ciphering methods developed to confuse FBI and NSA sniffing programs during Operation Sundevil (and it indeed succeeded in confounded them). It went on to be popularly used in the forums of online games like Counter Strike.

But L337 is not merely a ciphering system. The powerful thing about L337 is its propensity to evolve so quickly, to stretch language to the limits in its strange creole of argot, acronym, backronym, slang, cipher and hieroglyphica. I remember as a young player on internet game forums watching a language evolve before my very eyes as users pushed the envelope of orthography as far as it could go. Words and phrases decayed into seemingly random stream of glyphs that, against the odds, made sense.

Here’s two examples of L337 Speak:

omgomgomg j00 r teh sux0rz! i +0+411y pwnz j00!

()/\/\9()/\/\9()/\/\9 900 4|23 73|-| 5|_|>0|25 | +0+411] *\/\//\/3|) 900!

Both of which have the same meaning:

Oh my God [x3], you are the suckers! I totally owned you!

So far, so banally human….

It’s apt then that evasive and elusive 1337 speak demonstrates exactly how artificial intelligence is impossible. It’s worth mentioning here that the definition of AI varies from context to context, so I’ll take this moment just fix it at the point most common in popular perception: the ability of a machine to think and interact with humans without humans being able to tell if it’s a machine - i.e what we now call “strong AI” or "full" artificial intelligence. This is the kind of AI that poses a significant threat to human life as we know it.

Now, to put my case forward, I’ll need to turn to the philosopher Ludwig Wittgenstein, who thought about AI a good decade before the term was coined and Alan Turing came up with the imitation game that we know as the Turing Test. Wittgenstein wondered on paper if machines could ever think and came to the conclusion that they couldn’t. One such reason is that machines couldn’t possibly share the human "life form" that’s required for a shared horizon of meaning.

You see, human language is not a system of reference. Wittgenstein had thought that as a young man; he considered language as being like a picture of the world where words were referents and sentences were the statements of facts. But what’s the fact behind giving somebody the finger, or offering a high-five? So instead, Wittgenstein started to conceive of language not as a unitary, cohesive picture of the world, but as an infinite number of game-like activities with no unifying essence. Giving somebody the finger, winking, asking for something, counting things or lowering a flag to half-mast are ways of doing language in the infinite ways possible.

This is why games are the perfect analogy: games are infinitely variable, yet we can agree pretty quickly on what constitutes a game while it could have nothing in common with another. Solitaire has barely anything in common with baseball, Street Fighter II has nothing in common with playing fetch with a dog, but we know they’re games, right?

Why? Because, Wittgenstein argues, they have a family resemblance: games have no one thing in common, some games may have nothing in common, but they are connected by a pool of attributes that make up the sum of games. Solitaire involves cards just like top trumps, and top trumps has scores like baseball, and baseball is about defeating an opponent like kick boxing, which involves fighting just like Street Fighter II does. Solitaire and Street Fighter II have only one thing in common: they are games.

“Language games”, Wittgenstein holds, are similar, and importantly, just like games, they can be and are made up on the spot, that’s why L337 speak could evolve so fast yet still be understandable - it’s a matter of a shared horizon of experience, not merely understanding and following rules, but shaping them too: the intuition of the interpretation of rules with a mix of prescription and new precedent (a bit like laws in court). “If a lion could speak,” Wittgenstein famously stated, “we would not understand it.” Why? Because a lion is embedded in a different form of life: the lion’s form of life. Even if the lion could speak, the way it understands the world would be so inconceivably different from our own as a species that we wouldn’t understand it. The same would go for a computer. Here’s another one of Wittgenstein’s opaque aphorisms:

“Understanding a sentence, means understanding a language.”

The problem for AI is that language is more than a sum of its parts. Another philosopher, John Searle, used the now famous Chinese Room thought experiment to demonstrate that while AI could follow rules it wouldn’t be cognizant of them. I won’t go into that now, but you can read about it here.

L337 language then, in being conceived of to confound the algorithms, demonstrates why the natural use of language will never be attainable to machines.

What the hell does L337 have to do with Marketing? You ask. Well, our business is natural language, it’s understanding what makes people tick, intuition is our medium: we wink and nudge, we push the limits of language to get a rapport with people to build trust, empathy and understanding. Our jobs are safe. Or I should say, 0()r j0%z R 54Ph3.