Skip to Content
Artificial intelligence

Humans are still better than AI at Dota 2—for now

August 23, 2018

A team of professional video gamers prevailed over a group of cutting-edge AI algorithms in a landmark galactic mêlée held last night.

The game, which took place at the International Dota 2 Championships in Vancouver, Canada, showed that humans are still a lot smarter than AI software at complex tasks. The algorithms have previously defeated several teams of less-skilled players.

Five stars: The AI algorithms were developed by OpenAI, an AI research nonprofit funded by Elon Musk, Sam Altman, and others. Dubbed the OpenAI Five, each algorithm is a neural network that learns to play the game through what’s known as reinforcement learning.

Battle plan: Dota 2 involves two teams battling for territory within a sprawling space-themed battle arena. The game is challenging for AI programs because it is complex, requires teamwork, and unfolds over time (making it tricky to gauge successful progress). During last night’s match, the OpenAI Five exhibited some weird behavior before suffering a narrow defeat. Overall, it acquitted itself well against some of the world's best human players.

Cheat codes: Some have noted that, instead of relying on a view of the screen as a human player does, the OpenAI Five use the game’s application programming interface (API). These means the algorithms can instantly access more information about a game. Still, that shouldn’t detract from what is an impressive achievement for AI, despite the loss.  

Deep Dive

Artificial intelligence

A Roomba recorded a woman on the toilet. How did screenshots end up on Facebook?

Robot vacuum companies say your images are safe, but a sprawling global supply chain for data from our devices creates risk.

The viral AI avatar app Lensa undressed me—without my consent

My avatars were cartoonishly pornified, while my male colleagues got to be astronauts, explorers, and inventors.

Roomba testers feel misled after intimate images ended up on Facebook

An MIT Technology Review investigation recently revealed how images of a minor and a tester on the toilet ended up on social media. iRobot said it had consent to collect this kind of data from inside homes—but participants say otherwise.

How to spot AI-generated text

The internet is increasingly awash with text written by AI software. We need new tools to detect it.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.