|AI Goes Open Source To The Tune Of $1 Billion|
|Written by MIke James|
|Monday, 14 December 2015|
AI is a hot topic and it is clear that it could be an engine for great good or great evil depending on who or what controls it. OpenAI is a new non-profit that aims to put the control back in the people's hands.
The big headlines are that Elon Musk, Reid Hoffman, Peter Thiel, and Amazon Web Services are providing the funds. Given Elon Musk's public pronouncements on the possible dangers of AI this seems like a reasonable step. Some reports suggest that Google is also part of the team and this does not seem to be the case. In fact the three big players in the commercial AI world, Google, Microsoft and Facebook, are all missing.
The full list of funders is:
"Sam Altman, Greg Brockman, Elon Musk, Reid Hoffman, Jessica Livingston, Peter Thiel, Amazon Web Services (AWS), Infosys, and YC Research "
The funders have committed themselves to $1 billion, although this seems to be something more like a contingency fund than money up front. OpenAI states that it expects to spend only a tiny fraction of this in the next few years.
OpenAI's research director is Ilya Sutskever, who will be known to anyone following the rapid development of neural networks. As a research student of Geoffrey Hinton, Sutskever became one of the experts that Google acquired when it bought DNNresearch Inc, see Deep Learning Researchers To Work For Google. Now after three years with Google's Brain Team he has moved on to run his own research organization.
At the moment it isn't possible to be very clear about what sort of research will be top of the agenda:
"Our goal is to advance digital intelligence in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return."
It is clear that some of their concerns are going to be ethical:
"It's hard to fathom how much human-level AI could benefit society, and it's equally hard to imagine how much it could damage society if built or used incorrectly."
The idea seems to be to allow researchers to do work without having to keep their ideas locked up in companies or in patents:
"Researchers will be strongly encouraged to publish their work, whether as papers, blog posts, or code, and our patents (if any) will be shared with the world. "
This sounds great, but to an extent this is too little too late. Google has already patented some fundamental aspects of the modern approach to AI and hasn't said what it plans to do with them. For any hope of success OpenAI needs to get at least Google on board, if not Microsoft, Facebook and any other big company that pioneers AI in the future. This seems unlikely.
As well as renowned co-chairs Sam Altman and Elon Musk, OpenAI has some very well known advisers - Pieter Abbeel, Yoshua Bengio, Alan Kay, Sergey Levine, and Vishal Sikka. However, overall the group sounds a little top heavy with worthy top brass and not really sufficient people to do the spade work - perhaps this will change as research gets underway.
There is an initial research team of seven capable AI people - a tiny fraction of what Google, Facebook and Microsoft can throw at the problem - but OpenAI is hiring:
"We're slowly building a tight-knit team and thoughtful, open culture. We believe that unreasonably great results can come from a small group working in concert...
...We're based in San Francisco. We are hiring for intern and full-time positions."
Given the shortage of AI researchers, and the competition, they might find it hard to get the talent they need.
This is certainly a development to keep your eye on, but for AI the genie is out of the bottle and its going to take a lot more than a distant promise of $1 billion to put it back in.
or email your comment to: firstname.lastname@example.org
|Last Updated ( Monday, 14 December 2015 )|