The AI made a funny

When Watson won first place on Jeopardy, people worried. When Deep Blue beat Garry Kasparov, tech-minded communities were shook. When DeepMind’s AlphaGo beat a Go world champ, there were all sorts of rumours and rampant speculation. When OpenAI built a bot that taught itself to play DOTA 2 and beat the best in the world, people lost their minds. Today, you can go on Reddit—or read legacy publications—and witness the scare-mongering that occurs when the topic of AI, automation and the supposed impoverishment of human meaning is discussed. But I’m not worried about any of that. Let me explain why.

In his Paris Review interview, Umberto Eco considers humour. He says that we have yet to figure it out. Sure, theories exist. But Eco didn’t come across one which had satisfactory weight. He had his own suspicions regarding the existence of humour, and they centred around the observation that humanity is one of the only species that is conscious of its own mortality, and thus, humour is a coping mechanism that counters our knowledge of the inevitability of death. 

It’s an elegant hypothesis, and right now, I have neither the energy nor the capacity to explore its virtues and limitations. But the hypothesis—and Eco’s observation about the nature of humour— does highlight the trickiness, the complexity, of the human mind. Which is why I’m not worried about our ability to develop machine learning algorithms that can beat humans at simple and complex games. It’ll take more than that to compel me to act.

In fact, I’ll start my AI-related Doomsday preparations if—not necessarily when—we create an AI comedian.