Month: July 2019

Enlarge (credit: Blizzard)

Google's DeepMind is making its state-of-the-art AI play StarCraft II again. We previously saw the AI, called "AlphaStar," take on pro StarCraft II players in show matches, but now AlphaStar is ready to take on the public and faceroll scrubs on the 1v1 European multiplayer ladder.

Just like last time, AlphaStar is being built with the cooperation of Blizzard (StarCraft II's developer), and the official SC2 website has the details of AlphaStar's new incarnation. The in-game UI now has a "DeepMind opt-in button" on the 1v1 Versus menu, which will mix instances of AlphaStar into the human pool of multiplayer players. AlphaStar will be playing the 1v1 ladder anonymously, so you won't know if you're playing AlphaStar or a human (I mean, I guess you could try asking your opponent). Blizzard says that "having AlphaStar play anonymously helps ensure that it is a controlled test, so that the experimental versions of the agent experience gameplay as close to a normal 1v1 ladder match as possible." Players will be paired against AlphaStar according to the normal matchmaking rules, and a win or loss will count just as it would against a human.

The post contains a number of implementation details for this new version of AlphaStar, which sounds like a huge improvement over the version that played StarCraft II pros back in January. First, a number of improvements were made to bring the AI's speed capabilities more in line with a human player. As an AI company, DeepMind's stated goal with this experiment is to play SC2 on a level playing field and teach an AI things like thinking and long-term planning—basically, strategy. At a very high level, you could say the two big components to any StarCraft victory are "speed" and "strategy." DeepMind's previous AI experiments were in turn-based games like Chess and Go, where the speed at which you can move the pieces doesn't really matter. As a real-time game, speed is a big factor in any SC2 victory, and in previous games, AlphaStar at times demonstrated superhuman speed that gave it an unfair advantage and muddied the experiment results.

Enlarge / Coming October 25. And today, we know more about one cool, new multiplayer twist to the series. (credit: Activision / Infinity Ward)

LOS ANGELES—This spring, I visited game studio Infinity Ward to learn about its upcoming "reimagining" of Call of Duty: Modern Warfare. During the reveal event, the studio's reps mentioned multiple "pillars" of the upcoming CoD game, set to launch October 25, then focused on the game's single-player missions.

Shortly after that, someone walked into the room and said, "the rest of this event is indefinitely embargoed," then took the wraps off of arguably the best thing I've seen in a CoD game in years: a two-on-two "Gunfight" versus mode.

That embargo is finally up, following an official Gunfight gameplay reveal on Twitch this morning, so now I can finally tell you about a first-person versus mode that combines the original frantic action of 2007's CoD4:MW and the utter tension of Counter-Strike. In a gaming landscape forever altered by battle royale modes, Gunfight makes the case that sometimes, less is more.

Enlarge / The new recommendation engine is part of a new experimental Steam Labs branding.

For years now, Valve has been testing new approaches to filter the glut of Steam games down to the ones in which individual users are most likely to show an interest. To that end, the company is today rolling out a machine-learning-powered "Interactive Recommender" trained on "billions of play sessions" from the Steam user base.

In the past, Steam has relied largely on crowd-sourced metadata like user-provided tags, user-curated lists, aggregate review scores, and sales data to drive its recommendation algorithms. But the new Interactive Recommender is different, Valve says, because it works without any initial internal or external information about the games themselves (save for the release date). "Instead, the model learns about the games for itself during the training process," Valve says. "The model infers properties of games by learning what users do, not by looking at other extrinsic data."

Your own playtime history is a core part of this neural-network-driven model. The number of hours you put into each game in your library is compared with that of millions of other Steam users so the neural network can make "informed suggestions" about the kinds of games you might like. "The idea is that if players with broadly similar play habits to you also tend to play another game you haven't tried yet, then that game is likely to be a good recommendation for you," Valve writes.

Enlarge / The company does admit that it operates a website. (credit: Getty Images | NurPhoto )

Facebook is vigorously defending itself against a lawsuit brought by the attorney general of the District of Columbia—so vigorously, in fact, that it's basically claiming not to be a social network at all.

DC attorney general Karl Racine filed a lawsuit against the company in December, saying its actions related to the Cambridge Analytica scandal were in violation of DC consumer protection law. Facebook spent the last six months trying to have the lawsuit dismissed, but a federal judge in June rejected those requests and said the case should move forward.

As the case is now moving forward, Facebook submitted its response to the complaint (PDF) to the court this week, in which it denies "each and every allegation" in the suit.

The US is developing a new very-long-range air-to-air missile to counter Chinese and Russian weapons (transcript here). (video link)

It has been more than a decade since the US retired its last officially designated long-range air-to-air missile, the AIM-54 Phoenix—a Cold War-era weapon that could be used to take on multiple targets from more than 100 nautical miles (190km) at hypersonic speeds (above Mach 5). But the Phoenix was built for the F-14 Tomcat, and the longest-ranged weapon in the quiver of US Navy and Air Force fighter pilots since 2004 has been the AIM-120 Advanced Medium-Range Air-to-Air Missile (AMRAAM)—also known as the "Slammer"—with a somewhat shorter reach (160km, or 86 nautical miles).

Air superiority doctrine for the past two decades has focused on taking the enemy's aircraft out of the fight before they take off (with cruise missiles and stealth aircraft attacks on airfields) and technological superiority over whatever can get off the ground. The F-22 and F-35 both have emphasized stealth (or at least low observability) over weapon capacity and reach to ensure they can shoot down enemies before they're even seen. But that's not an equation that is balancing very well anymore as the US faces increased great-power competition from China and Russia. That is why the Department of Defense is rushing forward with development of a new long-range missile, the AIM-260—also known as the Joint Advanced Tactical Missile.