- QAnon-touting congressman sneaks ‘Epstein Didn’t Kill Himself’ into tweets Wednesday 7:12 PM
- Ocasio-Cortez met a famous drag queen–and the right melted down Wednesday 6:09 PM
- Woman says Lyft driver tried to kidnap her Wednesday 5:18 PM
- Debunking the right-wing conspiracy theories from today’s impeachment hearing Wednesday 4:29 PM
- Maroon 5 approves of the latest TikTok trend Wednesday 3:54 PM
- ‘One month left in the decade’ meme wants to know what you’ve accomplished Wednesday 3:53 PM
- Facebook Pay is the latest way to send your friends money Wednesday 3:31 PM
- Diving into ‘The Mandalorian’s first big shocker Wednesday 3:17 PM
- Disney+ will allow password sharing—to an extent Wednesday 1:12 PM
- Black server says manager refused to discipline coworkers who sent racist receipt Wednesday 12:47 PM
- Who is Jonah Hauer-King, Disney’s new Prince Eric? Wednesday 12:47 PM
- Cut Katherine Langford ‘Avengers: Endgame’ scene lands on Disney+ Wednesday 12:22 PM
- Planned Parenthood app to show abortion-seeking users their nearest options Wednesday 12:21 PM
- ‘The Imagineering Story’ offers touching insight into Walt Disney’s vision Wednesday 11:57 AM
- YouTube mom who was charged with child abuse dead at 48 Wednesday 11:39 AM
On Friday night, During Valve’s annual Dota 2 tournament, OpenAI presented players and fans with an AI bot that, in just two weeks, learned enough from itself and from Dota 2 pros that it’s begun to beat the legends. The bot then went onto defeat pro player Danylo “Dendi” Ishutin in a live 1-vs-1 match—twice.
Musk himself has lauded it as the first AI to defeat players in competitive esports.
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
To train the bot, OpenAI engineers said they matched the bot with a copy of itself, then turned it onto professional players. Ishutin himself remarked that the bot felt a little like a real person, but forfeited future matches with his opponent after such crushing losses. And though the bot’s demonstration was limited in gameplay, they’re hoping to extend its capabilities to compete in the full game in a 5-vs-5 match next year.
If you're not concerned about AI safety, you should be. Vastly more risk than North Korea. pic.twitter.com/2z0tiid0lc— Elon Musk (@elonmusk) August 12, 2017
Given Musk’s concern with AIs above North Korea, hopefully OpenAI’s bots might have wider capabilities by then.
H/T the Verge
Samantha Grasso is a former IRL staff writer for the Daily Dot with a reporting emphasis on immigration. Her work has appeared on Los Angeles Magazine, Death And Taxes, Revelist, Texts From Last Night, Austin Monthly, and she has previously contributed to Texas Monthly.