I recently had a conversation with a webdesigner about The Grid. The Grid is a website builder that uses ‘Artificial Intelligence’ to design websites according to the demands of the user and makes it look attractive. It will launch soon, and I’m considering to try it out. My conversation partner told me he believed human minds will always be necessary for this kind of thing. I replied that I wasn’t sure. One of the most striking films I recently saw on this topic was Her. In it, a program and a human become friends. I don’t want to spoil too much, but at some point the AI composes a jolly song. It’s fiction of course, but the story is self-explanatory and makes it credible. I bet it’s not the first time you hear that Artificial Intelligence is rising and taking over our jobs or even our lives, but have you ever really believed it? I’m starting to. If you would have asked a person before the war if a computer would ever be able to beat a human in chess, the answer would have been “No!”. Today, computers beat all champions. You could argue that chess is limited to the board and the predictable movements of the pieces, and therefore easy to calculate. Then, you could say that thoughts and words are unlimited, and that their sounds and meanings are too subtle for a computer to get, let alone to create with it. My answer would be: maybe. It might depend on how you program the AI. Let me take writing as an example. I’m not a grandmaster, but I’ve been doing it for a while now. A big part of it is technical: you attract attention with a title, build a structure of intro, middle and end, and try to choose your words such that they mean something. Build in some contrasts, break some grammatical rules. I don’t think people would disagree that the technical part is easy to learn for Artificial Intelligence. It’s the lived aspect that is harder. The part where emotions come in. Where meaning comes in. Where the sounds of the words dance around in your head. Where senses are triggered. The rhythm. Knowing what works and what doesn’t. Re-reading. Disagreeing with yourself. Making impossible choices. Creating symbols and metaphors. And yet when the text is done, there was only so much that a writer could do: the rest is what the reader creates for herself. If google can learn to recognize voices, can’t there also be a recognition of emotions in the tone of the voice? It’s all sounds, no? If facial recognition is possible, aren’t facial expressions the next step? With the increasing sensorial finesse of AI based systems, it could well be a matter of time before AI can discern a good wine from a bad one. Or a good story from a bad one. Give it control over the story, and it might improve it. I think another crucial thing to program is hunger. The insaturable need to take up information. To learn. The program should have limits, which force it to create. Digest, get stuff out. And it should be able to grow, but with a limited speed. Those are probably the hardest things to program, yet they have been attained with bacteria. Peristalsis, perhaps? I’m not an expert. Finally, to increase its status as a creator, the AI should have a drive to be acknowledged. If nowadays you can measure much of your societal recognition by the amount of views and likes of your webpage, then a ‘like = good – no like = bad’ algorithm should do the trick. Of course, you could further improve it with video information of people reading the words. Add up the factors and computers could become better at creating art, marketing themselves and being loved than humans ever have been.