Author: omohundro

Million AI Startups talk: AI and Games

On February 15, 2017, Steve Omohundro spoke to the Million AI Startups group about the opportunities in “AI and Games”:

Next Generation AI Games

Wednesday, Feb 15, 2017, 6:00 PM

Bootup Ventures
68 Willow Road Menlo Park, CA

70 Members Went

The use of Artificial Intelligence (AI) techniques in computerized games is as long as the history of AI itself. With recent advancements in AI, new possibilities are emerging for building video games that take entertainment to the next level. In these games every character can exhibit human-like intelligent behavior capable of incrementally learni…

Check out this Meetup →

Here is a pdf of the slides for the talk.

Video games are now a $100 billion industry. For comparison, global movie box office revenues for 2017 are estimated at $41.2 billion.

Blizzard’s “Overwatch” has generated $1 billion in revenue (from their Q1 2017 financial statement). It is their fastest growing franchise with 30 million registered players.

AI characters, like Cortana in Halo, are becoming more important to games. DeepMind and Blizzard are about to release a version of StarCraft II as an AI research tool.

There are at least 5 ways in which AI will improve games:

  1. AI as characters in games.
  2. AI as player of games.
  3. AI for improving VR/AR and game interfaces.
  4. AI for modelling learners and tuning games to their needs.
  5. AI for gamification of work and society.

New Voting Systems

Voting (and other forms of social decision making) are fundamental to our society. Today’s voting machines and technologies are antiquated, inefficient, and insecure. Here’s an excellent 8 minute description by Ron Rivest of how homomorphic encryption could help implement a better system:

Several groups are working to implement this kind of cryptographically secure voting on the blockchain:

In addition to better implementation technology, there are also a number of voting systems which are far superior to the one used in the US. Here’s a nice video describing the problems with “First Past the Post Voting”:

I’ve supported the “Center for Election Science” for years which is trying to institute Approval Voting (originally range voting). This is a simple modification to the current US system with much better properties:
More radical ideas are being explored in “Liquid Democracy” which allows voters to delegate their votes:
A somewhat more complex voting system “Quadratic Voting” is being hailed as one of the most significant advances in recent years. Here’s the paper:

Eric Posner says (

Glen Weyl has uploaded a new version of his paper, QuadraticVoting (written with Steven Lalley), to SSRN, which now includes the completed proofs. Quadratic voting is the most important idea for law and public policy that has emerged from economics in (at least) the last ten years.

Quadratic voting is a procedure that a group of people can use to jointly choose a collective good for themselves. Each person can buy votes for or against a proposal by paying into a fund the square of the number of votes that he or she buys. The money is then returned to voters on a per capita basis. Weyl and Lalley prove that the collective decision rapidly approximates efficiency as the number of voters increases. By contrast, no extant votingprocedure is efficient. Majority rule based on one-person-one-votenotoriously results in tyranny of the majority–a large number of people who care only a little about an outcome prevail over a minority that cares passionately, resulting in a reduction of aggregate welfare.

The applications to law and public policy are too numerous to count. In many areas of the law, we rely on highly imperfect votingsystems (corporate governancebankruptcy) that are inferior to quadratic voting. In other areas of the law, we require judges or bureaucrats to make valuations while knowing they are not in any position to do so (environmental regulation, eminent domain). Quadratic voting can be used to supply better valuations that aggregate private information of dispersed multitudes. But the most important setting is democracy itself. An incredibly complicated system of institutional self-checking (separation of powers, federalism) and judicially enforced constitutional rights try to correct for the defects of one-person-one-vote, but do so very badly. Can quadratic voting do better? Glen and I argue that it can.

And here are Tyler Cowen’s thoughts on it:
Interestingly, it’s been discovered that bees have been using this mechanism for millions of years to choose their next hive location! The energy bees spend on dances grows quadratically in proportion to the attractiveness of the site they saw.

Humans are doing democracy wrong. Bees are doing it right

There is a system that accounts for intensity of passion as well as idle opinion – hives have used it successfully for millions of years

AIBrain Talk: AI and Human Safety

On August 17, 2016 Steve Omohundro spoke to the “Million AI Startups” group about “AI and Human Safety”:

Top 10 AI Applications

Wednesday, Aug 17, 2016, 6:00 PM

AIBrain Inc
5 Palo Alto Square 1st Floor, CA

90 Members Went

Is AI flourishing now for every one?Can we make money out of AI? If so, how?In this vein, we are happy that the four presenters will lead the discussion in an effort to search for top 10 killer AI applications.AI and Human Safety, Steve Omohundro, Ph.D., President, Self Aware Systems AI and robotics will create $50 trillion of value over the ne…

Check out this Meetup →

AI and robotics will create $50 trillion of value over the next 10 years according to McKinsey. This is causing their rapid development but six recent events show the need to be careful as they are integrated into human society. In the past few weeks we’ve seen three Tesla autopilot crashes, the Dallas police using a robot to kill a suspect, a Stanford Shopping Center security robot running over a small child, and the first “Decentralized Autonomous Organization” losing $56 million due to a bug in a smart contract. As we move forward with these technologies, we will need to incorporate human values and new principles of security so that their human benefits can be fully realized.

Here is a pdf file of the slides.

TEDX Talk: What’s Happening With Artificial Intelligence?

The TED conference, started in 1984, has become the standard bearer for hosting insightful talks on a variety of important subjects. They have made videos of over 1,900 of these talks freely available online and they have been watched more than a billion times! In 2009 they extended the concept to “TEDx Talks”in the same format but hosted by independent organizations all over the world.

On January 6, 2016 Mountain View High School hosted a TEDx event on the theme of “Next Generation: What Will It Look Like?”. They invited both students from the school and external speakers to present. I spoke on “What’s Happening With Artificial Intelligence?”. A video of the talk is available here:


and the slides are available here:


I talked about the multi-billion dollar investments in AI and robotics being made by all the top technology companies and the 50 trillion dollars of value they are expected to create over the next 10 years. The human brain has 86 billion neurons wired up according to the “connectome”. In 1957 Frank Rosenblatt created a teachable artificial neuron called a “Perceptron”. Three-layer networks of artificial neurons were common in 1986 and much more complex “Deep Learning Neural Networks” were being studied by 2007. These networks started winning a variety of AI competitions besting other approaches and often beating human performance. These systems are starting to have a big effect on robot manufacturing, self-driving cars, drones, and other emerging technologies. Deep learning systems which create images, music, and sentences are rapidly becoming more common. There are safety issues but several institutes are now working to address the problems. There are many sources of excellent free resources for learning and the future looks very bright!

Eileen Clegg did wonderful real time visual representations of the talks as they were being given. Here is her drawing of my talk:


Edge Essay: Deep Learning, Semantics, and Society

Each year Edge, the online “Reality Club”, asks a number of thinkers a question and they publish the short essay answers. This year the question was “What do you consider the most interesting recent scientific news? What makes it important?” The responses are here:

My own essay on “Deep Learning, Semantics, And Society” is here:

VLAB Talk: AI, Deep Learning, and the Future of Business

On December 8, 2015 Steve Omohundro will be the special guest speaker at the VLAB Annual Holiday Party speaking about “AI, Deep Learning, and the Future of Business”. Followed by the “Chocolate Heads Movement Band”! See you there!

Here are the slides:

VLAB – AI, Deep Learning, and the Future of Business


AI Nexus Talk: Semantics, Deep Learning, and the Transformation of Business

On Saturday, November 28, 2015 at 2:00PM (Santiago, Chile time) Steve Omohundro will speak (remotely) at the Exosphere event “AI Nexus” on:

2:00 PMRemote Speaker: Steve Omohundro – Semantics, Deep Learning and the Transformation of Business

A pdf of the slides is here:

Chile – Semantics, Deep Learning, and the Transformation of Business

The SlideShare version is here:

Steve Omohundro, recognised Artificial Intelligence scholar, explains why semantics matter when talking about AI, what the deep learning trend is, and how business is going to be transformed by it.12182528_1002546619788560_500037707203850069_o

McKinsey predicts that AI and robotics will create $50 trillion of value over the next 10 years. Many predict that the recent technology of “deep learning” will be a big part of the transformation. Over 250 deep learning startup companies have attracted more than $1 billion of venture investment in the past year. Deep learning systems have recently broken records in speech recognition, image recognition, image captioning, translation, drug discovery and other tasks. Why is this happening now and how is it likely to play out? We review the development of AI and the pendulum swings between the “neats” and the “scruffies”. We describe traditional approaches to semantics through logics and grammars and the new deep learning vector semantics. We relate it to Roger Shepard’s cognitive geometry and the structure of biological networks. We also describe limitations of deep learning for safety and regulation. We show how it fits into the rational agent framework and discuss what the next steps may be.