December 7, 2019 News Magazine

This 16-game arcade for AIs tests their playing prowess

Figuring out just what an AI is good at is one of the hardest thing about understanding them. To help determine this, OpenAI has designed a set of games that can help researchers tell whether their machine learning agent is actually learning basic skills or, what is equally likely, has figured out how to rig the system in its favor.

It’s one of those aspects of AI research that never fails to delight: the ways an agent will bend or break the rules in its endeavors to appear good at whatever the researchers are asking it to do. Cheating may be thinking outside the box, but it isn’t always welcome, and one way to check is to change the rules a bit and see if the system breaks down.

What the agent actually learned can be determined by seeing if those “skills” can be applied when it’s put into new circumstances where only some of its knowledge is relevant.

For instance, say you want to learn if an AI has learned to play a Mario-like game where it travels right and jumps over obstacles. You could switch things around so it has to walk left; you could change the order of the obstacles; or you could change the game entirely and have monsters appear that the AI has to shoot while it travels right instead.

If the agent has really learned something about playing a game like this, it should be able to pick up the modified versions of the game much quicker than something entirely new. This is called “generalizing” — applying existing knowledge to a new set of circumstances — and humans do it constantly.

OpenAI researchers have encountered this many times in their research, and in order to test generalizable AI knowledge at a basic level, they’ve designed a sort of AI arcade where an agent has to prove its mettle in a variety of games with varying overlap of gameplay concepts.

The 16 game environments they designed are similar to games we know and love, like Pac-Man, Super Mario Bros., Asteroids, and so on. The difference is the environments have been build from the ground up towards AI play, with simplified controls, rewards, and graphics.

Each taxes an AI’s abilities in a different way. For instance in one game there may be no penalty for sitting still and observing the game environment for a few seconds, while in others it may place the agent in danger. In some the AI must explore the environment, in others it may be focused on a single big boss spaceship. But they’re all made to be unmistakably different games, not unlike (though obviously a bit different from) what you might find available for an Atari or NES console.

Here’s the full list, as seen in the gif below from top to bottom, left to right:

  • Ninja: Climb a tower while avoiding bombs or destroying them with throwing stars.
  • Coinrun: Get the coin at the right side of the level while avoiding traps and monsters.
  • Plunder: Fire cannonballs from the bottom of the screen to hit enemy ships and avoid friendlies.
  • Caveflyer: Navigate caves using Asteroids-style controls, shooting enemies and avoiding obstacles.
  • Jumper: Open-world platformer with a double-jumping rabbit and compass pointing towards the goal.
  • Miner: Dig through dirt to get diamonds and boulders that obey Atari-era gravity rules.
  • Maze: Navigate randomly generated mazes of various sizes.
  • Bigfish: Eat smaller fish than you to become the bigger fish, while avoiding a similar fate.
  • Chaser: Like Pac-Man, eat the dots and use power pellets strategically to eat enemies.
  • Starpilot: Gradius-like shmup focused on dodging and quick elimination of enemy ships.
  • Bossfight: 1 on 1 battle with a boss ship with randomly selected attacks and replenishing shields.
  • Heist: Navigate a maze with colored locks and corresponding keys.
  • Fruitbot: Ascend through levels while collecting fruit and avoiding non-fruit.
  • Dodgeball: Move around a room without touching walls, hitting others with balls and avoiding getting hit.
  • Climber: Climb a series of platforms collecting stars along the way and avoiding monsters.
  • Leaper: Frogger-type lane-crossing game with cars, logs, etc.

You can imagine that an AI might be created that excels at the grid-based ones like Heist, Maze, and Chaser, but loses the track in Jumper, Coinrun, and Bossfight. Just like a human — because there are different skills involved in each. But there are shared ones as well: understanding that the player character and moving objects may have consequences, or that certain areas of the play area are inaccessible. An AI that can generalize and adapt quickly will learn to dominate all these games in a shorter time than one that doesn’t generalize well.

The set of games and methods for observing and rating agent performance in them is called the ProcGen benchmark, since the environments and enemy placements in the games are procedurally generated. You can read more about them, or learn to build your own little AI arcade, at the project’s GitHub page.


Source: TechCrunch

Tags: in Uncategorized
Banner
Related Posts

Fresh off a $530M round, Aurora acquires lidar startup Blackmore

May 23, 2019

May 23, 2019

Aurora, the self-driving car startup backed by Sequoia Capital and Amazon, is in an acquiring mood. The company, founded in...

Incognito mode comes to Google Maps

May 7, 2019

May 7, 2019

Being able to browse the web without leaving the usual local traces is a valuable tool for a lot of...

With warshipping, hackers ship their exploits directly to their target’s mail room

August 7, 2019

August 7, 2019

Why break into a company’s network when you can just walk right in — literally? Gone could be the days...

Elon Musk finally hosted meme review with the co-creator of Rick and Morty

February 22, 2019

February 22, 2019

Tesla and SpaceX CEO Elon Musk has been teasing — and his fanbase has been making pleas — to host...

Terramera raises $45 million for its technology to reduce the use of chemicals in agriculture

September 5, 2019

September 5, 2019

Terramera, a Canadian company selling bio-pesticides and seed treatments to reduce the use of chemicals in agriculture, has raised $45...

Finally, an official Craigslist app

December 4, 2019

December 4, 2019

Fancy websites and services come and go, but Craigslist endures. And now one of its main shortcomings is fixed: there’s...

The winners of The Europas Awards 2019 display Europe’s continuing diversity and ambition

July 1, 2019

July 1, 2019

The Europas Awards for European Tech Startups came around again last week (Thursday 27th June), and once again proved that...

This robot learns its two-handed moves from human dexterity

May 29, 2019

May 29, 2019

If robots are really to help us out around the house or care for our injured and elderly, they’re going...

Todoist releases major update to simplify task management

October 22, 2019

October 22, 2019

Bootstrapped tech company Doist, the company behind popular task management Todoist, is releasing a major update called Todoist Foundations —...

Travel activities platform Klook raises $225M led by SoftBank’s Vision Fund

April 9, 2019

April 9, 2019

We recently noted that SoftBank’s Vision Fund has stepped up its deal-making in Asia this year, and today it added...

Qualcomm launches its next-gen 5G modem and mmWave antenna

February 19, 2019

February 19, 2019

There can be little doubt that 5G is going to take center stage at this year’s Mobile World Congress in...

True Balance raises $23M to bring its payments app to more small cities and towns in India

October 15, 2019

October 15, 2019

South Korean startup True Balance, which operates an eponymous financial services app aimed at tens of millions of users in...

T-Mobile quietly reported a sharp rise in police demands for cell tower data

July 12, 2019

July 12, 2019

T-Mobile has reported a small decline in the number of government data requests it receives, according to its latest transparency...

iPhone 11 Pro hands-on

September 10, 2019

September 10, 2019

More than any other iPhone event in recent memory, today’s big launch was content-first. Apple began the show with several...

Apple puts accessibility features front and center

June 9, 2019

June 9, 2019

Although the meat of Apple’s accessibility news from WWDC has been covered, there still are other items announced that have...

Comments
Leave a Reply

Your email address will not be published. Required fields are marked *