OpenAI today announced the launch of Spinning Up, a program designed to teach everyone deep reinforcement learning.
OpenAI is known to create funk-looking agents in virtual environments where self-contained learning such as Humanoid v2 is taught or POLO collaborates with the University of Washington.
In enhancing learning, reward signals are given to an agent in an environment that is stimulated to maximize the reward to reach a goal.
RL played a role Key AI breakthroughs like AlphaGo by Google DeepMind and agents trained in environments like Dota 2.
The program is not only launched to show people how reinforcement learning works, but to achieve progress toward OpenAI's overall goal of creating Artificial Generic Intelligence (AGI) by involving more people from other areas than of computer science.
"The AI security solution requires people with a wide range of expertise and perspectives, and many relevant professions have no connection to engineering or computer science. However, all participants need to learn enough about the technology to make informed decisions, and several parts of Spinning Up need to address that need, "Spinning Up's documentary text says.
Spinning Up has been created through working with OpenAI fellows and scholarship holders – people with a curious interest in AI, but little experience in the field.
The program was not only geared towards the goal of secure delivery of AGI, but also because there is not a single textbook today, The knowledge is currently included in a number of different lectures and influential research.
"Our package here is intended as a missing mid-step for people who are excited by Deep RL and would like to learn Use it or make an example But you have no clear idea of what to study or how to turn algorithms into code. We tried to make this starting point as helpful as possible, "the project says.
The Spin Up initiative is part of a larger set of educational services that OpenAI wants to make available, a spin-up workshop will be held in February 2019 and another workshop will be held in conjunction with the Center for Human Compatible AI (CHAI) at the University of California, Berkeley.