Something is growing exponentially into something that is of great significance.


Recently I have noticed a great deal of news on some of the world’s biggest names in technology have open sourced their artificial intelligence (AI) systems. It’s been happening over a span of just a few months in this bizarre domino effect. I get this sense that something is growing exponentially into something that is of great significance. So I want to know what that something is, why is this happening and what does it mean. I’m writing about it in the hope of getting a true understanding.


A Brief Timeline — The Domino Effect

The titans of Silicon Valley in over the past few months have made some of the most sophisticated AI programs available to anyone. Here is a brief timeline of which major companies have recently open sourced their AI technologies:

  • June last year Amazon open sourced ‘Alexa’ it’s voice command response system inhabiting the company’s Echo device.1
  • November last year Google open sourced the software library for ‘TensorFlow’ which is the perceptual and language program, it’s seen as the heart of its image search technology. 
  • A month later in December last year Facebook released the designs of the GPUs for its AI algorithm system Big Sur. Big Sur is the custom hardware designs that run Facebook’s M personal assistant.
  • A day after Facebook open sourced its AI hardware, OpenAI was launched. OpenAI is a non-profit AI research company that is backed by Elon Musk and Sam Altman. The company is dedicated to “advancing digital intelligence in the way that is most likely to benefit humanity as a whole…as broadly and evenly distributed as possible.” It’s vowing to share everything it builds.
  • By the end of the second week of the new year, Chinese Web company Baidu released its key AI software WARP-CTC on GitHub. WARP-CTC was created to improve speech recognition in Baidu’s end-to-end speech recognition program Deep Speech 2.
  • Only weeks apart towards the end of January, Microsoft followed suit and released it’s AI toolkit CNTK (Computational Network Toolkit) on GitHub. Like Baidu, this software also focuses on deep learning and speech recognition.
  • Last month in July, Microsoft released AI system Project Malmo which is the system used to test out AI algorithms using the Minecraft world.
  • Only a couple of days ago, OpenAI released its first toolkit called ‘GYM’. The toolkit is aimed at helping solving the most difficult problem in machine learning path which is unsupervised learning. ‘OpenAI Gym is used for comparing reinforcement learning algorithms and it allows researchers to test their algorithms by having them play games, control a robot simulation, or complete tasks.’6


Reasons for Open Sourcing

It seems unusual that companies would reveal the methods of their core businesses. Why release trade secrets, wouldn’t that undermine these companies’ competitive edges? Well sharing in a sense is a way of competing. By openly sharing software and hardware designs, this can accelerate the progress of AI which ultimately advances these companies own interests. Within the tech industry there is this common view that technology companies like Google, Facebook and Amazon are not in the businesses one might assume. Google and Facebook are not really in the business of selling ads and Amazon is not in the business of selling merchandise. They are actually powered by your data, their currency is users. In order to draw and keep users to its products they need to quickly innovate and produce better and more products. These companies open source their AI software because they wish to be the foundations on which other people innovate. AI is central because by design, it learns, adapts and even makes decisions. AI is more than a product, it’s a product generator.’1 In the future, AI will be used to identify and capitalise on new opportunities by innovating new products. Open sourcing AI allows these companies to stay on the cutting edge of technology and helps pave their way to their future.

There is irony in that, for companies to have competitive advantage they need to give away their trade secrets to their competitors and to the community. Open sourcing to a large community of researchers can help accelerate and foster new breakthroughs in AI. “There is a network effect. The platform becomes better as more people use it,” says Yann LeCun, a founding father of deep learning, who now oversees AI work at Facebook. “The more people that rally to a particular platform or standard, the better it becomes — the more people contribute. It’s how they share their thoughts and ideas. In the closed source world, developers don’t have a lot of room to move.”7 Open sourcing also is a way of recruiting serious engineering talent and retaining talent. In the field of deep learning in particular, the community of researchers who excel at deep learning is relatively small. “It is certainly a competitive advantage when it comes to hiring researchers,” says Sam Altman (president of Y Combinator).2 Another reason for open sourcing is it can help reduce the cost of machines. Software and knowledge about algorithms are non-rival goods. Making them freely available would enable more people to use them, at a low marginal cost.8 The more companies start using the designs, manufacturers can build the machines at a lower cost.

Accelerating the evolution of AI has large positive outcomes as well as negative outcomes. There is this belief the more people that look at the code, the more likely bugs and potential negative outcomes can be worked out. Open sourcing of code allows many people to think through the consequences both individually and together. This is the foundation of OpenAI, the non-profit AI research company associated with Elon Musk. Most people have seen or read about Elon Musk and other tech giants concerns over AI and how it could be our biggest existential threat.9

There really isn’t any surprise for Musk’s motive for forming OpenAI. Musk and other top leaders of Silicon Valley such as Peter Thiel and Y Combinator founded the organisation on the premise that any advances in AI should benefit humanity as a whole, “build value for everyone, not just for shareholders”, and be completely transparent to the world. Co-chair Sam Altman reveals, "We think the best way AI can develop is if it’s about individual empowerment and making humans better, and made freely available to everyone, not a single entity that is a million times more powerful than any human. Because we are not a for-profit company, like a Google, we can focus not on trying to enrich our shareholders, but what we believe is the actual best thing for the future of humanity."10

It seems like a win win situation, OpenAI is enormously beneficial for the wider community and it’s also beneficial for Musk and Altman. Musk runs Telsa which are building self-driving cars, which can benefit from deep learning technology in enormous ways. Musk can also create a far more powerful pool of data that can help feed the work. ‘Altman says that Y Combinator companies will share their data with OpenAI, and that’s no small thing. Pair their data with Tesla’s, and you start to rival Google — at least in some ways.’2 There maybe altruistic and selfish motives here but it doesn’t diminish the value of OpenAI.


Image source: backchannel.com

Criticism Against

The ideology of OpenAI is based on the premise that if everyone has access to the capabilities of intelligent machines for their own purposes then no one has the unnecessary advantage over others. There is some criticism against this. Steven Levy a technology journalist brings up an excellent question in an interview with both co-chairs Musk and Altman on Backchannel. Levy asked whether their plan to freely share this technology would actually empower bad actors by giving state of the art AI to the Dr. Evils of this world. “If I’m Dr. Evil and I use it, won’t you be empowering me?”10 Musk and Altman responded by stating that they felt the power of many will outweigh the power of a few. “Just like humans protect against Dr. Evil by the fact that most humans are good, and the collective force of humanity can contain the bad elements,” said Altman, “we think its far more likely that many, many AIs, will work to stop the occasional bad actors.”11 Although, not everyone in the field agrees with this. Nick Bostrom an Oxford philosopher points out that if you share research without restriction, bad actors could grab it before anyone has ensured that it’s safe. Bostrom says “If you have a button that could do bad things to the world, you don’t want to give it to everyone.”12

Another criticism to open sourcing is that faster development and deployment of AI, running at breakneck speeds to win the AI race would make people less inclined to put safety precautions in place. This view is shared by both Bostrom and Miles Brundage an AI Policy Research Fellow at Oxford.Their criticism seems valid and may hold some weight especially in a culture that embraces this ‘fail fast’ mentality.

There is also another criticism that this rapid development of AI will lead to the birth of ‘a super-intelligence that recursively improves itself, that will eventually reach an escape velocity and become orders of magnitude smarter than any human.’Bostrom in his book ‘Superintelligence: Paths, Dangers, Strategies’ argues that if machine brains surpass human brains in general intelligence, then this new super intelligence could replace humans as the dominant lifeform on Earth.13 Open sourcing AI will lead to the shortening of the path to this super-intelligence.

Image source: chuckmanrobots.blogspot.com.au
Although digital visionary Kevin Kelly argues that this view that intelligence is single dimensional of IQ is fundamentally wrong and describes that our own intelligence is a symphony of cognitive notes and instruments.14 We have different kinds of intelligence like deductive, spatial reasoning and emotional intelligence. There are hundreds of different varieties of types of notes that make up our intelligence. AI is going to be very similar in that there will be combinations of different intelligence. We cannot put AIs on a linear map because it’s going to be multi-dimensional. “It’s not going to be this intelligence explosion of AIs bootstrapping themselves into higher orders.”14

Kelly describes the origins of technology as relating to Darwin’s theory of evolution and the big bang. He defines evolution as more than a philosophy it’s a theory where there is a cosmic force. This cosmic force does not have a single direction, it has this outward radiation of many different dimensions. The directions of evolution and technology are not converging to a specific omega point, they are radiating outward in many directions.14 Echoes of Darwin’s beautiful and elegant theory of evolution still continues to reverberate everywhere.

“The Darwinian revolution is unfinished, like relativity, the theory of evolution should be considered a ‘special theory, applicable only in the limited sphere of biological change. It needs to be extended to a ‘general theory’, which also applies to the human world” expressed by science writer Matt Ridley in his book ‘The Evolution of Everything.’15

Image source: evolutionposters.com


Thoughts and the Profoundness

Here are my personal thoughts that I have arrived at in this point in time, I’m sure these thoughts will change and evolve as I continue on this journey for more knowledge. I believe that the Googles and the Facebooks are already rapidly pushing AI, it seems as though that it is inevitable that AI is going to continuously advance whether we like it or not. Technological development of AI will continue to advance however, I believe that leaving the advancement of AI to the major players may result in outcomes that may not be as humanistic. Although Google has it’s ‘no evil mantra’ and Mark Zuckerburg is not focusing on earnings, I believe having a non-profit organisation like OpenAI to act as a counterbalance and keep the big tech companies in check is overall beneficial. If there’s this constant prominence on safety and AI being nudged in the direction with the greater good for humanity in mind, then hopefully this will steer AI development in a good direction. By embracing and engaging in AI technology we can steer it, we can have much more of a chance of optimising the benefits and minimising the harm. Prohibition of the sharing of AI technology is not the answer. Open sourcing endorses democratisation and transparency. I believe that the democratisation and transparency of AI technology and information in general is very powerful and significant. It encourages more efficient and informed decision making, it also fights against corruption and levels out the playing field. More importantly, it drives honesty, cultivates accountability and instills trust.

This democratised access to AI technology and resources in today’s sharing and creation economy is very profound as it is the foundation of this oncoming technological innovation explosion that is unfolding. Kelly describes “Cognification and Artificial Intelligence is by far the most profound and altering thing. I suspect it’s going to produce change on the level of the industrial revolution and go beyond touching every aspect of our lives.”14 Many other visionaries share this idea that this technological revolution has very profound impacts. Economist and political advisor Jeremy Rifkin describes it greatly in his books ‘The Empathic Civilization’16 and ‘The Third Industrial Revolution’.17 Andrew McAfee and Erik Brynjolfsson both academics at the MIT Center for Digital Business also liken this incredible boom in technologies to that of The Industrial Revolution. “The Industrial Revolution ushered in humanity’s first machine age — the first time our progress was driven primarily by technological innovation — and it was the most profound time of transformation our world has even seen”. They describe that we are now entering another technological revolution, ‘the second machine age,’18 Albert Wenger venture capitalist and partner at Union Square Ventures also mentions in his interview on the Singularity podcast that “A change is a foot that is deeply profound as when we went from bringing agrarians to the industrial society. Everything will change, it will have to change.”19 History tells us that technological innovation has lead to great progress to humanity and I believe it is accelerating human evolution. I like Albert Wenger’s world view that ‘knowledge is the foundational human project’.19 My view is I think the exploration and discovery of new realms of meaning is what progresses human evolution and ‘knowledge’ is key to that. If we want to keep progressing and evolving we need to embrace exploration and discovery in new technologies.

With going forward to embracing and engaging in this technological revolution there needs to be this balance between optimism and realism. There seems to be this resurgence of this kind of unbridled optimism about the possibilities of the future and how we are creating a world of increasing abundance. Engineer and entrepreneur Peter Diamandis constantly conveys this techno-inspired optimist view that we are in this radical new era of increasing abundance in his podcasts and book ‘Abundance’.22 Matt Ridley in his book ‘Rational Optimist’ debunks apocalyptic thinking 20, cognitive scientist Steven Pinker asserts that violence has been decreasing through history in his book ‘The Better Angels of Our Nature.’21 Although I have this feeling that excessive optimism can sometimes be a negative and can lead to complacency. It can cause us to look at the world through rose tinted glasses consequently blinding us from potential harms and leads us to being unprepared to threats. Although I also see the other side of the coin where too much pessimism can be damaging, it can lead to the feeling of powerlessness and then to inaction. Noam Chomsky, the founder of modern linguistics, one of the founders of the field of cognitive science, states:

“We have two choices. We can be pessimistic, give up and help ensure that the worst will happen. Or we can be optimistic, grasp the opportunities that surely exist and maybe help make the world a better place. Not much of a choice.”23 

I want to be optimistic about the future but instill realism into that optimism, be mindful of the true problems that lie ahead and recognise the duality that this technological revolution will create. Technological progress makes the world better but it also brings new challenges which we should acknowledge and hold with the highest regard.




Bibliography

1) Shafto, Patrick. “Why Big Tech Companies Are Open-sourcing Their AI Systems.” The Conversation. N.p., 22 Feb. 2016. Web. 11 Aug. 2016. http://theconversation.com/why-big-tech-companies-are-open-sourcing-their-ai-systems-54437/.

2) “Can Open AI Project Protect Humanity from Future’s Skynets?” Daily Star Albany. California Turkish Times Daily News & Headlines, n.d. Web. 11 Aug. 2016. http://www.albanydailystar.com/science/can-open-ai-project-protect-humanity-from-futures-skynets-mesquite-daily-news-14116.html/

3) Novet, Jordan. “Baidu Open-sources Its WARP-CTC Artificial Intelligence software.” VentureBeat. N.p., 14 Jan. 2016. Web. 11 Aug. 2016. http://venturebeat.com/2016/01/14/baidu-open-sources-its-warp-ctc-artificial-intelligence-software/

4) Furness, Dyllan. “Top Tech Companies Open Source Their AI “Secrets”.”TechEmergence. N.p., 02 Feb. 2016. Web. 11 Aug. 2016. http://techemergence.com/top-tech-companies-open-source-their-ai-secrets/

5) Osborne, Charlie. “Microsoft’s Project Malmo AI Platform Goes Open Source | ZDNet.” ZDNet. N.p., 8 July 2016. Web. 11 Aug. 2016. http://www.zdnet.com/article/microsofts-project-malmo-goes-open-source

6) Muoio, Danielle. “We Just Got Our First Glimpse of What Elon Musk’s AI Company Is Working on.” Tech Insider. Business Insider, 27 Apr. 2016. Web. 11 Aug. 2016. http://www.techinsider.io/elon-musks-openai-released-its-first-toolkit-2016-4

7) Metz, Cade. “Facebook Open Sources Its AI Hardware as It Races Google.”Wired.com. Conde Nast Digital, 12 Oct. 2015. Web. 11 Aug. 2016. http://www.wired.com/2015/12/facebook-open-source-ai-big-sur/

8) Bostrom, Nick. “Strategic Implications of Openness in AI Development.” (2016): 155–73. Nick Bostrom. 2016. Web. 11 Aug. 2016. http://www.nickbostrom.com/papers/openness.pdf

9) Muoio, Danielle. “Here’s What Elon Musk’s Secretive AI Company Is Working on.” Tech Insider. Business Insider, 12 Apr. 2016. Web. 11 Aug. 2016. http://www.techinsider.io/what-elon-musks-openai-is-working-on-2016-4

10) Levy, Steven. 2015. “How Elon Musk and Y Combinator Plan to Stop Computers From Taking Over.” Backchannel (blog), December 11. Available at: https://backchannel.com/how-elon-musk-and-y-combinator-plan-to-stop-computers-from-taking -over-17e0e27dd02a

11) Metz, Cade. “Elon Musk’s Billion-Dollar AI Plan Is About Far More Than Saving the World.” Wired.com. Conde Nast Digital, 15 Dec. 2015. Web. 11 Aug. 2016. http://www.wired.com/2015/12/elon-musks-billion-dollar-ai-plan-is-about-far-more-than-saving-the-world/

12) Metz, Cade. “Inside OpenAI, Elon Musk’s Wild Plan to Set Artificial Intelligence Free.” Wired.com. Conde Nast Digital, 27 Apr. 2016. Web. 11 Aug. 2016. http://www.wired.com/2016/04/openai-elon-musk-sam-altman-plan-to-set-artificial-intelligence-free/

13) Bostrom, Nick. Superintelligence: Paths, Dangers, Strategies. UK: Oxford UP, 2014. Print.

14) Kevin Kelly on The Inevitable: 12 Forces That Will Shape Our Future. Dir. Nikola Danaylov. Perf. Kevin Kelly and Nikola Danaylov. Singularity Weblog. Nikola Danaylov, 25 July 2016. Web. 11 Aug. 2016. https://www.singularityweblog.com/kevin-kelly-on-the-inevitable-12-technological-forces-that-will-shape-our-future/

15) Ridley, Matt. The Evolution of Everything: How New Ideas Emerge. N.p.: HarperCollins, 2015. Print.

16) Rifkin, Jeremy. The Empathic Civilization: The Race to Global Consciousness in a World in Crisis. New York: J.P. Tarcher/Penguin, 2009. Print.

17) Rifkin, Jeremy. The Third Industrial Revolution: How Lateral Power Is Transforming Energy, the Economy, and the World. New York: Palgrave Macmillan, 2011. Print.

18) Brynjolfsson, Erik, and Andrew McAfee. The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies. New York: W. W. Norton, 2014. Print.

19) VC Albert Wenger on Basic Income and World After Capital. Dir. Nikola Danaylov. Perf. Albert Wenger and Nikola Danaylov. Singularity Weblog. Nikola Danaylov, 17 June 2016. Web. 11 Aug. 2016. https://www.singularityweblog.com/albert-wenger-world-after-capital/

20) Ridley, Matt. The Rational Optimist: How Prosperity Evolves. New York: Harper, 2010. Print.

21) Pinker, Steven. The Better Angels of Our Nature: Why Violence Has Declined. New York: Viking, 2011. Print.

22) Diamandis, Peter. “Peter Diamandis.” Peter Diamandis. N.p., 2016. Web. 11 Aug. 2016. http://www.diamandis.com/

23) Stafforini, Pablo. “Why I Choose Optimism Over Despair.” Chomsky.info. Roam Agency, 14 Feb. 2016. Web. 11 Aug. 2016. https://chomsky.info/why-i-choose-optimism-over-despair/

24) Wallpaper Zone. N.p., n.d. Web. 12 Aug. 2016. http://wallpaper.zone/img/4042141.jpg
© Christine Calo 2016. Please do not reproduce without the expressed consent of Christine Calo. Powered by Blogger.