Get perfect grades by consistently using www.customizedassignments.com. Place your order and get a quality paper today. Take advantage of our current 20% discount by using the coupon code GET20
Order a Similar Paper Order a Different Paper
Save your time - order a paper!
Get your paper written from scratch within the tight deadline. Our service is a reliable solution to all your troubles. Place an order on any task and we will take care of it. You won’t have to worry about the quality and deadlinesOrder Paper Now
DeepMind AlphaStar (Starcraft II bot) links
Last Thursday (1/24), Google’s DeepMind team announced a bot, AlphaStar, which they claimed showed high-level human play on Starcraft II for the first time. Some discussion ensued: what does this show, and what does it mean for AI & Games, or AI in general? Since it’s breaking news quite relevant to our course, we’ll spend some time this week covering both the results and some of the discussions and controversies.
I’ve collected some relevant links below:
DeepMind StarCraft II Demonstration video
- Commented recording of five human-vs-machine games
- There’s an explanation of the learning algorithm at 1:16:30
- Explanation of the algorithms and bot training league
- Some interactive visualizations of strategies
- Downloadable data
Reaction blog posts and articles:
- DeepMind AI thrashes human professionals at video game StarCraft II – New Scientist
- My first impressions of AlphaStar – David Churchill
- An Analysis On How Deepmind’s Starcraft 2 AI’s Superhuman Speed is Probably a Band-Aid Fix For The Limitations of Imitation Learning – Aleksi Pietikäinen
- What follows AlphaStar for Academic AI Researchers? – Ben Weber
- DeepMind AlphaStar Analysis and Impressions (video)
DeepMind announced a claim to have reached top human-level performance (with some fine print) on StarCraft II with their bot AlphaStar, the first time that’s been done for a real-time strategy game. I posted a collection of primary-source and early reaction links; see “DeepMind AlphaStar (Starcraft II bot) links” in under Content in Blackboard.
Your assignment is to dig a little bit into these materials and come up with a preliminary opinion.
1. Overall, what do you think this shows? Is it a major advance for AI? A minor one? Not sure? Does anything in particular stand out at you from the videos, DeepMind blog post, and following debate as being particularly noteworthy?
2. If you had the ability to ask DeepMind to do follow-up experiments or further data analysis, what would you want to see?
(Expected length: A page or two, i.e. 1-2 paragraphs per question.) 12 times new roman
Books you can look at: Chapter 1-2-3-4 of Playing Smartby by Julian Togelius,
This class is about AI games theory.