Honestly, I think the title from the New York Times as "the father of AI" is a bit presumptuous. I'm not an expert by any means in machine learning or artificial intelligence, but I do know a fair amount about computer history.<p>Potential "father of modern AI" - even that is stretching a bit!<p>The fact is, machine learning and artificial intelligence is a story in history of fits and starts; of springs and winters; of successes and failures.<p>If anyone could be called such, the fathers of AI would belong to Warren McCulloch and Walter Pitts in 1943, who came up with the model for an artificial neuron:<p><a href="https://en.wikipedia.org/wiki/Artificial_neuron" rel="nofollow">https://en.wikipedia.org/wiki/Artificial_neuron</a><p>From there, it was people standing on each other's shoulders, building ever upward and outward. These steps and systems have ultimately led to today's deep learning networks.<p>Other machine learning techniques trace to various techniques and methods in statistics and probability theory; then you have the whole arena of computer/machine vision research.<p>Right now, we're in the midst of yet another AI/ML "spring", after a fairly long "winter". Mr. Schmidhuber can certainly claim a bit of status as being one of many who help institute a thaw leading to today - but he is by no means alone (I'd argue that one of the earliest for today's thaw might be Lecun).
> "Juergen Schmidhuber, often referred to as the father of AI"<p>What?<p>reading further...<p>> he New York Times recently referred to him as a would-be father of AI.<p>Clicking the link to the NYTimes article<p>> When A.I. Matures, It May Call Jürgen Schmidhuber ‘Dad’<p>English as a second language person here, but is it just me or the Bloomberg subtitle does not reflect at all what the NYT title is saying?
This is a sloppy piece making grand claims about a man who likes to make grand claims about himself. By most accounts, Schmidhuber is only marginally involved with nnaisense, lending his name to the endeavor so that some of his graduate students can raise money. They have no use case, no product and frankly no business sense, which is one reason why they only money backing them comes from an unknown investor based in Madrid.
Good to see. I had the opportunity to see Jurgen speak last year - he was excellent - entertaining, surprisingly deep and able to field any questions with humour and insight. He gets a lot of flak for being obsessed with being credited for his work, but I think he does it for historical accuracy rather than ego.<p>Research oriented AI startups are experiencing serious brain drain to large companies because of the money they can afford to pay, I hope this secures Nnaisense for a couple of years and they make the most of that time
I attended my first NIPS this year, and found Juergen to be a very engaging speaker, with the RNN symposium organized by him and his colleagues being my favorite part of the conference. A popular phrase that was being thrown around during the conference was "learning to learn" or "meta learning", with one of the papers even being titled "learning to learn by gradient descent by gradient descent"[1] Juergen seemed very passionate about the subject and he gave a cool talk around his Godel Machine[2], and sparked interesting conversation during the panel discussion. I wouldn't be surprised if "learning to learn" or "meta learning" replaces "deep learning" as the AI-word of 2017.<p>[1]<a href="https://papers.nips.cc/paper/6461-learning-to-learn-by-gradient-descent-by-gradient-descent.pdf" rel="nofollow">https://papers.nips.cc/paper/6461-learning-to-learn-by-gradi...</a><p>[2]<a href="https://en.wikipedia.org/wiki/G%C3%B6del_machine" rel="nofollow">https://en.wikipedia.org/wiki/G%C3%B6del_machine</a>
Interestingly enough, Googling "Father of AI" yields several results including John McCarthy, Marvin Minsky, and Alan Turing.<p>Google's provided results point at McCarthy while the first mention of Schmidhuber is at the 9th result from the New York Times.
For more details, refer to Schmidhuber's own website: <a href="http://people.idsia.ch/~juergen/" rel="nofollow">http://people.idsia.ch/~juergen/</a> and his excellent review paper, "Deep Learning in Neural Networks: An Overview": <a href="https://arxiv.org/pdf/1404.7828.pdf" rel="nofollow">https://arxiv.org/pdf/1404.7828.pdf</a>
Schmidhuber cracks some hilarious jokes when he is on stage.<p>"The other day I gave a talk and there was just a single person in the audience, a young lady. I said young lady it's very embarrassing but apparently today I'm going to give this talk just to you. She said okay but please hurry I gotta clean up here."
Hinton. Hands down. Just read his early work on deep nets. Unsupervised, no ensemble, nailed MNIST and self-awareness of features. And he has a great sense of humor.
There's a long line of succession... take your number, Jurgen: McCulloch, Pitts, Fukushima, Kohonen, Hopfield, Jordan, Elman, Werbos, LeCun, Hinton, Bengio... only to name very few.<p>You can also go even further... <a href="https://www.youtube.com/watch?v=laJX0txJc6M" rel="nofollow">https://www.youtube.com/watch?v=laJX0txJc6M</a>