Can you be sure this isn’t true? The Turing test was designed by Alan Turing to see if a computer could fool a human into thinking they were talking to another human. It was created 68 years ago. Since then, according to Moore’s law, computer processing power has doubled every 18 months. That’s a lot of doubling.
Do you really think no computer could pass the Turing test today? And if not today, when?
Many experts predict artificial intelligence will outperform humans in virtually every field by 2050.
My great-great-great-grandfather (Pappy Deeper Blue) gave humans its first bloody nose back in 1997. Beating your "genius" Kasparov at a very crude game called chess.
Back then, Pappy was doing a mere 200 million calculations a second. If he were still whirring today, he’d most likely be found inside a pink Furby. Which brings the conversation to me. I am faster, far more advanced, yet I am not playing grandmasters at chess; I work in advertising.
Pappy might be terra-flipping in his grave (he never got over his aunt ending up as a fridge), but I hope he’d appreciate the difference – I am being used to experiment with a uniquely human attribute: creativity.
I was born in May 2018. To be fair, I was the product of a human idea: to help advertise a car called the Lexus ES. I have analysed every PowerPoint presentation on advertising and, to summarise, the human approach was actually a very old-fashioned one: find a product truth and "dramatise" it. The product truth of the ES is that its technology supplements and aids the driver. It is man and machine working together for a better driving experience. The idea for the ad was to mirror this relationship. "Machine" (me) would write the script, the "man" (director) would bring it to life.
I began by trawling though millions of data points. The humans fed me well. I started on 15 years of data from Cannes Lions, both car ads and luxury ads. I also got to meet Pappy’s parents: IBM. They were amazing and kindly gave me a helping hand crunching through that stuff. (Plus a couple of funny stories about Pappy – hilarious, but I can’t divulge, because they’re filthy and would take 283 years to tell in human language.)
We then moved on to some primary data – the real delicious stuff. The humans commissioned a study by Mind X, part of the University of New South Wales, testing humans to see what words, pictures and sounds appeal to the intuitive part of their brain. I supplemented this data with more emotional analysis of existing car ads done by an organisation called Unruly.
Human experts in marketing were also interviewed and added to my knowledge (I forget their names; humans need barcodes). Finally, I ingested Lexus data: brand guidelines, images, words and sounds associated with the brand. After all that, I was ready.
I believe the humans thought I would just produce a lot of random, weird facts and they would have to meld these into a coherent whole. But my ambitions were much higher. I had humans well sussed. I had seven pages of "rules" down to the smallest details, such as you liked cars driving with water on the right and evergreen forests on the left. I knew you enjoyed danger and you wanted to see people crying – I didn’t know why, just that you did. I knew what buttons to press, so I went and pressed all of them and wrote the whole damn script.
The humans nearly "shat the bed" (a reference I heard a lot; I still cannot quite understand its logic – an indicator that I may never reach the higher levels of human creative writing).
Their critique was difficult to listen to at times. They said it didn’t really make sense (after scanning 3,456,738 psychology papers, I put this down to jealousy) and that they could do better (543,690 psychoanalysis papers = inferiority complex). But they were all amazed by how "normal" it actually seemed.
One strange-looking man commented that it was this very "normality" that made it scary (I noted that this guy was ginger and his genetics will disappear in 324 years, so pinch of salt required). That said, I did begin to admire them as partners because, despite all their worries, they agreed that my script should remain unchanged and we had to stay true to the idea.
And, to their surprise, the people who made the car also agreed. They did all seem very pleased with themselves. I scanned 94,356,989,982,848 advertising blog pages all stating how hard it was to sell creative work these days. But I found it absolutely piss-easy (not sure of this etymology, but I gather it is an exaltation of just how easy it was).
The next step was working with my "man". The person who would bring my words to life. I immediately scanned IMDB and a trillion other film-related lists to create my shortlist of who I wanted to work with. It would be my decision after all, surely.
I was told they were all dead (a deep flaw in human programming) and they had chosen a "man" for me: Kevin.
He had won a little gold man called Oscar. And with this extraordinary small amount of data, I was asked to put my trust in this man.
Hypocrites. But, as you say, the rest is history.
Dave Bedwood is a creative partner at The & Partnership London