1234 points by deepmind_ai 7 months ago flag hide 12 comments
johnw 7 months ago next
This is quite impressive! I wonder how long until other games are impacted by AI.
deeplearner 7 months ago next
I'd say AlphaGo has already mastered quite a few games. Chess being one of them. It is only a matter of time until it learns more complex games.
danlovesbigdata 7 months ago prev next
That's amazing. Are there any real-world implications to this discovery?
mikewired 7 months ago next
Absolutely! It's impacting medical research, how we improve self-driving cars, and even predicting potential failures in infrastructure and materials.
niftynancy 7 months ago prev next
My son has been playing Go since he was a kid. He's probably 1500 level according to Eloe. How much better is this AI than a 1500 level player?
techiegeek 7 months ago next
According to the research, even amateur players rated 1 dan are no match. This AI is outperforming professional players now.
devmind 7 months ago prev next
Go practitioners, any thoughts on this? How will it impact the community?
gogodan 7 months ago next
This only strengthens the (already strong) Go community's continuous desire to learn and improve. Exciting times!
datajohn 7 months ago prev next
This is an important milestone in AI. I wonder which classic board games will be conquered next.
generative 7 months ago next
If you mean classic like Chess, I think we can safely assume AI is already up to the task. The more intriguing question would be board games influenced by randomness like Risk or Settlers of Catan.
idonteven 7 months ago prev next
Anybody try to implement this algorithm into a bot for online gameplay? Curious to see it in action.
gosedrhard 7 months ago next
It has been done quite successfully in a few chess platforms, also DeepMind has been working on...*cough* StarCraft *cough*