From Papers to Robots: OpenAI is Making AI Research Cool Again

Image for post
Image for post

Last week, OpenAI announced one of its first practical applications. The non-profit disclosed that is reprogramming robots developed by Fetch Robotics with algorithms that allow them to train themselves by trial and error. The new robots used a technique called deep reinforcement learning that has been one of the first areas of focus of OpenAI.

When I first learned about OpenAI months ago, I thought it was a very interesting idea that wasn’t going to get passed the typical bureaucracy of non-profit organizations. At the end, most of the breakthroughs in software history have been designed with commercial purposes even is they followed open standards or open source distribution models. With AI being at the center of the efforts by powerhouses like Microsoft, Google, Facebook or IBM, how effective can OpenAI be? The answer? Very very effective.

In just a few months, OpenAI has released several products and some incredible research about AI disciplines. In addition to the work with Fetch Robotics, in April OpenAI launched the OpenAI Gym: a toolkit for measuring and comparing reinforcement learning algorithms. The AI non-profit has also published some incredibly insightful research on unsupervised learning and generative models. Those are some very impressive milestones for an eight-month-old company but is is enough to succeed? In the ultra-competitive AI market, the success of OpenAI might as much dependent on network effects than to the quality of its technology. Here are a few ideas that I think are going to be key for the success of OpenAI.

· Attracting top AI research talent: OpenAI must continue attracting top AI research talent that prefers to advance AI without the constraints of specific commercial goals.

· Partnerships with software and hardware companies: Similar to the partnership with Fetch Robotics, it will be key that more software and hardware companies start leveraging OpenAI tools nad frameworks in their products.

· Powering AI startups: Similar to the previous point, OpenAI research and software can become an incredible asset for AI startups that don’t have the resources of Google or Facebook to attract top AI research talent.

· Nurturing a developer community: Attracting developers and building an open source community around OpenAI products should be another area of focus for the OpenAI team.

These are just some ideas that might play a role in the success of OpenAI. Without a doubt, the non-profit is off to a fast start and has the potential of leveling the playing field between AI-powerhouses and the rest of the world.

Written by

CEO of IntoTheBlock, Chief Scientist at Invector Labs, Guest lecturer at Columbia University, Angel Investor, Author, Speaker.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store