Every week, my team at Invector Labs publishes a newsletter to track the most recent developments in AI research and technology. You can find this week’s issue below. You can sign up for it below. Please do so, our guys worked really hard on this:
From the Editor
Can artificial intelligence(AI) formulate complex strategies? The process of strategizing has long been considered a magic ability of the human mind. Formulating a strategy requires evaluating many options to predict outcomes in the future. Games has long been considered a great environment to evaluate the strategic abilities of AI systems. Among the popular games in the market, StarCraft II is widely recognized as one of the most complex in terms of its strategic capabilities. While AI systems have made progress in strategy games such as Atari, Mario, Quake III Arena Capture the Flag, and Dota 2, none have come close to challenge StarCraft II.
This week, Alphabet’s subsidiary DeepMind unveiled AlphaStar, an AI systems that was able to outperform human players in StarCraft II. AlphaStar leveraged advanced multi-agent reinforcement learning techniques to learn to play StarCraft II by playing against itself. The result? AlphaStar was able to outperform a team of professional players by winning all but one of the games it played. The techniques learned from AlphaStar can be applied to many other complex strategic problems such as scientific research or policy formulation. A great week for AI indeed!
Now let’s take a look at the core developments in AI research and technology this week:
DeepMind unveiled AlphaStar, a reinforcement learning agent able to master complex strategies in StarCraft II
Google released Natural Questions, a new large corpus for training question-answering systems.
IBM Research published a paper about the principles of ethically aligned AI
Cool Tech Releases
Facebook open sourced LASER, a library for zero-shot transfer of natural language processing(NLP) models.
The Facebook AI Research(FAIR) team released a predictive model for deciding where to place energy lines in emerging countries.
Microsoft announced a series of new releases to help companies adopt AI
AI in the Real World
Researchers from the Massachusetts Institute of Technology(MIT) created a model that can identify instances where AI models make dangerous errors.
The New York Times discusses the challenges that government faced in the world of AI.
Stephen A. Schwarzman is chairman, CEO and co-founder of Blackstone published an article in the Washington Post about the challenges of creating an ethical AI.