What kind of future do people have in a world full of smart machines? A possible answer to this question was given by chess legend Garry Kasparov at the WeAreDevelopers Congress in Berlin. His vision of the future is already a reality in some part.
“What is artificial intelligence (AI) anyway? The gateway to paradise or the gateway to hell”, Garry Kasparov asked in Berlin on 6. of June. “None of this – AI is just a technology.” However, one that will change our society greatly – as was the case with chess.
Kasparov first played against a chess computer in 1985 – 32 games simultaneously. He won them all. “The machine was weak and bulky back then,” he said yesterday in Berlin. That’s changed. “Current chess computers don’t play perfectly yet, but they make almost no mistakes.” You can not definitely say that about people, says Kasparov.
Deep Blue is stupid – and wins anyway
In 1996 IBM held a competition with the chess computer Deep Blue. Kasparov won 4:2, “but that was the turning point”. For the first time ever a reigning world chess champion lost in tournament mode against a machine. Just one year later Deep Blue was already too strong for Kasparov. The computer won 3.5 to 2.5. “Although it was about as intelligent as your alarm clock.”
In chess, the competition between man and machine is definitely over. “The machines have won,” said Kasparov, “in the end, they always win.” So are we heading for the end of the mankind? People should now concentrate on uniting their imagination and creativity with the machines. “Then it’s not just the machines that win – we all win,” Kasparov said.
Robots vs. toddlers
Pablo Picasso once said that computers are absolutely useless. They could only give answers, but not ask questions. “The problem is that AIs don’t know which issues are really relevant to humanity,” Kasparov said. Although the machines had improved, they were still only successful in very specific areas.
Besides, machines still can’t move well. “You might be able to keep up with a 2-year-old kid today.” This is not much different from the Boston Dynamics robots either. Although they would look great, they could only solve very specific problems.
Why AIs Are Not Creative
Kasparov stressed that technologies are not good or bad, but fundamentally neutral. “The big question is how people use the technologies they create in their everyday lives.” It is not surprising that Microsoft’s Twitter bot has cut out racist tweets, for example. After all, there are many racist people, and AIs would only reflect the mistakes and biases of people.
Machines have the potential to make people smarter. This can be seen, for example, in health care, where AIs can support doctors. In addition, people who use machines are more efficient than people without machines. That’s logical, Kasparov said. Machines, in contrast to humans, could calculate probabilities very well. The human being trumps for it with creativity, which AIs cannot offer yet.
It’s better together
At the end of his presentation Kasparov presented the following conclusion:
- Technologies that we now call intelligent were bad at first, then suddenly useful, and in the end sometimes even disruptive.
- People and machines can make processes more efficient if they work closely together.
- Machines will be able to solve big problems. But it is still the people who ask the right questions and have to define what success is.
- It is becoming easier and easier to use smart tools. That is why less and less training and retraining will be needed in the future.
“Machines will not replace us,” Kasparov concluded. “They’ll get us ahead!” (“We aren’t being replaced. We’re being promoted!”)
This article first appeared in Netzwoche