# Deep-Learning the Hardest Go Problem in the World

DevFeed: [Deep-Learning the Hardest Go Problem in the World](<https://devfeed.tech/articles/deep-learning-the-hardest-go-problem-in-the-world-20163.md>)

Original publisher: [Read original article](<https://blog.janestreet.com/deep-learning-the-hardest-go-problem-in-the-world/>)

Author: David Wu

Published: 2019-12-06T00:00:00Z

Content type: article

Language: en

Sources: [Jane Street](<https://devfeed.tech/sources/jane-street.md>)

Topics: [AI Research](<https://devfeed.tech/topics/ai-research.md>), [Deep learning](<https://devfeed.tech/topics/deep-learning.md>), [Bot](<https://devfeed.tech/topics/bot.md>), [Replication](<https://devfeed.tech/topics/replication.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [ai-research](<https://devfeed.tech/tags/ai-research.md>), [bots](<https://devfeed.tech/tags/bots.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [games](<https://devfeed.tech/tags/games.md>), [go](<https://devfeed.tech/tags/go.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [replication](<https://devfeed.tech/tags/replication.md>)

## AI overview

This article reports updates to KataGo, including a 19-day self-play run using up to 28 V100 GPUs that reached the strength of ELF OpenGo in roughly one-fiftieth of the computation. It then examines Igo Hatsuyoron problem 120, widely regarded as one of the deepest and hardest Go problems.

## Source excerpt

Updates and a New Run