Review: The State of Artificial Intelligence in a Nutshell

On Thursday evening the 28th of September a former neuroscience graduate returned to the lecture halls of ETH Zurich. Pascal Kaufmann, now founder of Starmind International AG, gave a keynote speech about artificial intelligence in connection with the "A.I.: Connecting the Dots" event.

At the beginning of his speech, Pascal Kaufmann gave the floor to a man who knows how to thrill people with words: a video of a speech about the Apollo-mission John F. Kennedy held in 1962 at Rice University, which summarized the atmosphere of departure during this time. The moon landing inspired an optimism about technology and an enthusiasm about the future. The same spirit Kaufmann hopes to generate in our generation. He wants Switzerland to take a leading role in creating a movement to decode the human mind itself and create true A.I. He believes in the power of ideas and in outstanding talents. That's why Kaufmann has held this speech, to talk about his views on A.I., the state and progress of robotics and the hype that is created around these topics. He wants to discuss three points: Beware of the robot hype, beware of the AI hype, and let us unite to tackle the mind code and create A.I.

The A.I. and the robot hype - about expectations and reality

After these inspiring words, Pascal Kaufmann shows the audience two pictures of the same event eight years apart. Both show the election of the pope in Rome – but with a big difference. In the latter picture, from 2013, millions of phone displays illuminate the scene. The two pictures show one thing:  digitalization has changed our daily life, technology made a huge step forward. If we believe in what the media says, so has the science around A.I. But science-fiction movies or exaggerated reports in the news make it hard to judge: what is true about A.I.?

Because of that, robotics in 2017 has a problem: the expectations in robots and A.I. are (too) high. This becomes obvious with two videos. Scenes from famous Hollywood blockbusters like "I, Robot" in comparison with videos of real robots – properly accompanied by circus music – show the huge gap between expectations and reality. We expect robots with perfect functionality, personalities and a human-like look – like we know them from the movies. Real robots however, show little functionality and look more like toys with some malfunctions.

Of course, there are service and social robots we benefit from, and lots of money is invested in companies that design these kinds of robots. But Pascal Kaufmann is convinced that not so many robots will survive, because they look so similar and are not much more than "an iPhone on wheels". That's why he wants to make sure that we consider the following to "prevent the bubble from bursting": Don't contribute to the A.I. and the robot hype. When scientists and especially the media talk about robots or A.I., they need to be realistic and honest. Because the expectations they create need to be fulfilled.

A.I. – just human intelligence put into source code?

With another famous man Pascal Kaufmann returns to the main theme of this keynote speech – A.I.
A picture shows Napoleon Bonaparte playing chess against a robot. It was the famous "Turk" who made everyone believe that a machine was able to beat humans in a game of chess. The A.I. hype already started 200 years ago! People believed that they created true A.I. It took them 84 years to find out, that real chess players were sitting underneath the Turk, playing against the opponents. Kaufmann draws a line from this historical trick to today's form of A.I. He criticizes that most of A.I. is really just "cheating": human intelligence put into source code.

Kaufmann is looking for another way of A.I. We shouldn't try to copy the human brain – a task almost impossible to fulfill given the complexity of brain neurons – but should rather focus on the principle of nature.

The future of A.I.

Do we play in team robot or in team human? Pascal Kaufmann is sure: we need to play in team human, because humans are linear and technology is exponential. He wants us to use newest technology like the HoloLens and think about a world were jobs are executed by robots.

With a quote from Geoffrey E. Hinton, a cognitive psychologist and computer scientist, Kaufmann addresses the students in the audience. Hinton said that the future of A.I. depends on a student who is deeply suspicious of everything he has heard and that they need to start all over. In the same way Switzerland, ETH Zurich and especially the students need to take action regarding A.I.

Kaufmann started his speech with a Greek god: Apollo, who gave his name to the human spaceflight program of NASA. Reaching the end of his speech, he reconnects A.I. with another Greek god: Prometheus. The one who brought fire to mankind. Will we be able to "bring fire" into our robots, to build true A.I. rather than putting human intelligence into source code? Pascal Kaufmann is sure that it is just a matter of time until we will.


With the kind support of the cogitofoundation and ETH Global

Cogito Foundation

ETH Zürich


A.I. - Team (