TEDxDelft 2012 | Performer: Bennie Mols – The one thing computers and robots will never be able to do…

written by Rob Speekenbrink, 5 September 2012, filed as article

When he was eight years old, Bennie Mols (1969) wanted to become a professor. He failed. Instead, he became a science journalist, science writer and science communicator. He holds master degrees in physics (TUE) and philosophy (UvA), as well as a Ph.D.-degree in physics (TUD, 1999). He never stops wondering about science in general and physics, computer science, mathematics, neuroscience and psychology in particular. And he tries to make a living by sharing the fruits of his wondering with a broad audience.

Bennie Mols wrote more than four hundred popular science-articles for, among others, NRC Handelsblad, NWT Magazine, KIJK, Intermediair and De Ingenieur. He published several books and has participated as a speaker in more than two hundred radio broadcasts explaining science to a broad audience. Earlier this year he published his latest popular-science book: β€˜Turings Tango βˆ’ Waarom de mens de computer de baas blijft’, about the quest to build thinking machines. The book combines the latests insights in artificial intelligence with those from the brain- and cognition sciences and philosophy. Turings Tango also honors the birth of computer pioneer Alan Turing exactly one hundred years ago.

Mols loves artificial intelligence, but argues that artificial intelligence will always remain different from human intelligence (and that there is nothing wrong with that). It’s the biology stupid!

1 comment

One Response to TEDxDelft 2012 | Performer: Bennie Mols – The one thing computers and robots will never be able to do…

  1. Jay says:

    I have just read Bennie Mols’ article in the latest New Scientist.

    His remark that most computer scientists (which ones?) disagree with Kurzweil’s vision really stood out for me. It was pretty much the only statement made in the article which was unsupported by rational argument.

    I wanted to know more about who Bennie Mols was. So I visited his blog at benniemols.blogspot.com.

    I quickly stumbled across a post where Bennie Mols literally writes “Kurzweil’s Hollywood-like magic with Moore’s Law”.

    It’s this sort of irrational Kurzweil hatred that keeps him from coming in full touch with reality. After all, this is nothing more than defensive pessimism.

    Defensive pessimism is nothing more than an inefficient ego protection mechanism. It certainly doesn’t quality as a way of getting in touch with reality. How can it? Defense pessimism is by definition not equal to rationality.

    Bennie Mols has an emotional stake in the idea that humans will always be AI’s masters. But this is by no means a given.

    I certainly *do* agree that it would be much safer to meld human intelligence with artificial intelligence. Building AI from scratch is a direct existential threat. Upgrading our own intelligence would be a much less threatening strategy.

    Kurzweil has written on the mind/machine merger extensively. But Bennie Mols conveniently forgets all about this when he goes off into a frenzy against Kurzweil.

    A human mind / machine merger would, eventually, certainly lead to the first human being that upgrades him/herself to such an extent, that he/she transcends the human condition by miles. At this point, this person might as well have been the result from a superior AI built from scratch. He/she will be a superior intelligence. Whether it be a natural or artificial intelligence… is besides the point.

    No matter how you look at it… there will come a day when there will exist an intelligence beyond that of the most intelligent human beings currently in existence.

    There is no point in trying to convince one self that human intelligence will always be the best of the best. AI is creeping its way up the cognitive ladder. First chess, then driving cars. Who knows what’s next. But there will always be those who keep telling themselves AI will never get up to their level, let alone beyond it.

    Bennie Mols’ argument seems to be that Moore’s Law is already ending. Moore’s Law only says something about the number of transistors on a chip, and absolutely nothing about the exponential acceleration in computing. So that’s quite a miss already.

    He completely ignores the 10 page quantum computer special in the very same edition of New Scientist as the one in which his article appeared. He also ignores new technologies that can easily keep exponential growth in computing going, such as nanotube computing, photonics computing and 3D circuitry stacking.

    Kurzweil has definitely written about the evolution of computing power. We are currently on the fifth computing platform. Every time a computing platform runs out of steam, a new platform is found and computing power keeps growing exponentially. Kurzweil has the data for this and it goes back decades. It’s quite a mountain of evidence that supports an ongoing exponential trend.

    When Kurzweil critics such as Bennie Mols make unproven statements about the end of exponential growth in computing (which is what they mean when they say Moore’s Law is dying), it reeks of researcher’s pessimism.

    The article above flat out states that Bennie Mols has failed to become what he wanted. That’s the expected result of contraining oneself with limiting beliefs.

    Kurzweil, on the other hand, is now working for Google where he works on a very ambitious AI project. Not exactly the kind of results you’d get from limiting oneself in any way.

    Let’s contemplate this for a few seconds. Kurzweil has been hired by Google, the very company that pushed AI up a level by building a self driving car. Does this position lend Kurzweil no credibility whatsoever?

    Sure… most computer scientists disagree with Kurzweil. But none of those computer scientists seem to work at Google.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>