Artificial intelligence took a step forward last night, at an annual tournament for players of the tactical wargame Defense of the Ancients 2. A bot created by the Elon Musk-backed nonprofit OpenAI defeated champion human player Danylo “Dendi” Ishutin in two back to back demonstration matches.
OpenAI first ever to defeat world's best players in competitive eSports. Vastly more complex than traditional board games like chess & Go.
— Elon Musk (@elonmusk) August 12, 2017
Defense of the Ancients 2 – commonly referred to as DOTA 2 – is what’s known as a multiplayer online battle arena, or MOBA. Players control one of dozens of different characters with varying abilities, and compete to collect items and control territory. It’s currently one of the most popular games from Valve, the publisher that organized last night’s event, and one of the most popular competitive e-sports games worldwide.
Get Data Sheet, Fortune’s technology newsletter.
AI developers have recently shown that computers can dominate the best human players in Go and chess. But DOTA 2 has far more variables and possible board states than even Go, meaning decision-making is much more complex. The game also takes place in real time rather than discrete turns.
The bot’s victory, then, adds up to more than just fun and games. OpenAI describes it as “a step towards building AI systems which accomplish well-defined goals in messy, complicated situations involving real humans.” That includes applications like delivery routing, strategic planning, and traffic management.
According to The Verge, last night’s demonstration did reduce some of the game’s complexity. Perhaps most significantly, while Ishutin was defeated in a 1-on-1 match, DOTA 2 is normally played by opposing teams of five players each. OpenAI says it plans to continue developing its software so it can play full-scale matches.
It might seem odd that Elon Musk would sponsor AI development at all, since he’s been vocal about the threat he thinks the technology poses to humanity. But OpenAI is aimed at building ‘safe’ AI and “influencing the conditions under which AI is created” – potentially by helping Musk push for greater regulation of the technology.