Future of work
Machines, if you can’t beat them, join them
When Kasparov was beaten by Deep Blue in chess, I have to admit that I was rooting for the computer. I wanted Kasparov to win, but I was happy I wasn’t the only one that couldn’t beat the computer. You see my own intellect or rather lack thereof had fallen to a much less powerful computer years before. I got revenge since I was playing on a cracked game. I adjusted the level of computer hints to infinite. Then played the computer with its help and starting winning again.
At the time though, there was widespread concern that the end of human intelligence as the dominant force in the world had come to an end. The reality was that Deep Blue sat on top of a database of all the prior human moves and was programmed by humans. In fact, afterwards Kasparov quipped that he could have beaten Deep Blue if he had a standard laptop. Kasparov went on to suggest that tournaments should allow players to use laptops. This suggestion has come to pass in “freestyle” chess tournaments with any combination of humans, computers and human computer teams. The Winner? A human and computer team of amateurs, who picked the best moves from three different computer programs’ suggestions.
Kasparov remarked, “Human strategic guidance combined with the tactical acuity of a computer was overwhelming.”
At Greycroft, we have started to invest against this theme. Our first investment is in WorkFusion, which does for data collection and enrichment work what the assembly line did for manufacturing. WorkFusion is a software-as-a-service platform that pairs machine learning with human data analysts to train automation. It’s like an IronMan suit for knowledge workers. Users are getting 10x improvements in terms of throughput and cost, but aslo have dramaticaly different scalabliity, speed and accuracy. The company already has 14 of the 17 largest financial information providers as customers and are rapidly transforming how these companies think about their work processes to create the information products they sell. The broader question is how many problem sets will be solved. If it’s as broad as we think this will revolutionize many industries. Computers were invented to solve a small set of problems, but when they became general purpose machines they changed everything.
The best defense is to get yourself ‘smart’ with computers. Stephen Hawking also shares these concerns and suggests that humans have to augment their intellect with computers. A suggestion he has taken to heart. He uses a device that converts his brain waves into speech.
For me, I wholeheartedly believe in this strategy and have AngelList and big data strategies like Correlation Ventures breathing down my neck. In fact, we use a number of data products to augment our process and have our own proprietary data that we mine. Now, I need to go and figure out how to hack the GPU in my PlayStation to do pattern recognition. Anyone have a Torx screwdriver?
So in the end the combination of computers and humans are really unbeatable and will continue to be for a long time. At least until the computers decided the best cure for spam is to kill all the humans as Elon Musk ponders or we all go on permanent holiday as the machines are doing all the work.
