DeepMind and Blizzard Open StarCraft II as an AI Research Environment • Hacker News Board

Re: DeepMind and Blizzard Open StarCraft II as an AI Research Environment

#281

Post by avar » Fri, Aug 11, 2017, 12:41 PM UTC

yannyu wrote: post #275

avar wrote: post #253

Earlier quoted context omitted.

You still end up with a computer that can perform 3 actions all within the first millisecond of a second and still end up with human-like 180 APM (3 APS), even though no human could replicate what it just did.

How is that any different than someone who goes 100MPH and then 20MPH and claims to have not violated a 65MPH rule because the average speed was 60MPH for the trip? The units of the measurement do not dictate how the measurement is made.

Because when the cops measure your speed they can do so in an arbitrarily small time period of their choosing, and inertia and power requirements mean you can't be going 1000 MPH one milliseconds and 10 MPH the next.

You could decide to measure APM by saying that the time difference between any two actions extrapolated to sustaining that rate over a minute or seconds couldn't exceed the APM or APS, but as I've explained such a measurement would unfairly give the human player an advantage because humans are capable of bursts they couldn't sustain over longer periods.

Re: DeepMind and Blizzard Open StarCraft II as an AI Research Environment

#282

Post by naveen99 » Sun, Aug 13, 2017, 9:40 AM UTC

Synaesthesia wrote: post #130

naveen99 wrote: post #86

> even strong baseline agents, such as A3C, cannot win a single game against even the easiest built-in AI. Then, why not release code for the built in ai, and improve on it ? Or is the built in ai cheating ?

The built in ai is scripted whereas they’re trying to teach agents to learn the game from scratch with some sort of reward-based/machine learning approach.

still, why not use machine learning to tweak the script instead of starting from scratch.

Re: DeepMind and Blizzard Open StarCraft II as an AI Research Environment

#283

Post by visarga » Sun, Aug 20, 2017, 2:51 AM UTC

bytefactory wrote: post #65

gwern wrote: post #55

Earlier quoted context omitted.

I think it is doable in under 5 years, but this critically depends on the resources invested by DM and other DL orgs. Deep RL is hugely demanding of computational resources to iterate your designs - for example, the first AlphaGo took something like 3 GPU-years to train it once (2 or 3 months parallelized); however, with much more iteration, DM was able to get Master's from-scratch training down to under 1 month. Now…

> (It might not even be as complex as people think ... Yeah, I suspect you're right. Eliezer was alluding to this with the AlphaGo victory as well: > ... Human neural intelligence is not that complicated and current algorithms are touching on keystone, foundational aspects of it. https://www.facebook.com/yudkowsky/posts/10153914357214228?p... I can't decide if I would be bummed or excited if that turns out to be the…

> rather than depending on those elegant, serendipitous breakthroughs that much of human progress has been built on

That's brute forcing as well. One such elegant idea comes every million(billion?) people. Random people would just output random ideas.