AI News, DeepMind: Homepage artificial intelligence

DeepMind cofounder Mustafa Suleyman leaves for Google

„After a wonderful decade at DeepMind, I’m very excited to announce that I’ll be joining @Kent_Walker, @JeffDean and the fantastic team at Google to work on opportunities &

Google and parent company Alphabet are increasingly incorporating DeepMind’s AI work into their other divisions, thanks in large part to Suleyman’s efforts, he said.

„Mustafa played a key role over the past decade helping to get DeepMind off the ground, and launched a series of innovative collaborations with Google to reduce energy consumption in data centres, improve Android battery performance, optimise Google Play, and find ways to improve the lives of patients, nurses and doctors alike,“

Two years ago, that division was found to have improperly obtained access to the health records of 1.6 million UK patients.

Google DeepMind gamifies memory with its latest AI work

Over time, you realize the mistake even before disaster strikes.  Likewise, you know over years when you made the wrong choice, like choosing to become a manager at Best Buy rather than a pro-ball player, the latter of which would have made you so much more fulfilled.  That second problem, how a sense of consequence develops over long stretches, is the subject of recent work by Google's DeepMind unit.

The authors of the paper, 'Optimizing agent behavior over long time scales by transporting value,' which was published November 19th in Nature Magazine's Nature Communications imprint, are Chia-Chun Hung, Timothy Lillicrap, Josh Abramson, Yan Wu, Mehdi Mirza, Federico Carnevale, Arun Ahuja, and Greg Wayne, all with Google's DeepMind unit.  The point of departure for the game is something called 'long-term credit assignment,' which is the ability of people to figure out the utility of some action they take now based on what may be the consequences of that action long into the future — the Best Buy manager-versus-athlete example.

And Allen Newell and Marvin Minsky, two luminaries of the first wave of AI, both explored it.  Of course, AI programs have a form of action-taking that is based on actions and consequences, called 'reinforcement learning,' but it has sever limitations, in particular, the fact it can't make correlations over long time scales the way it seems people are doing with long-term credit assignment.  'Humans and animals evidence behaviors that state-of-the-art (model-free) deep RL cannot yet simulate behaviorally,' write Hung and colleagues.

In particular, 'much behavior and learning takes place in the absence of immediate reward or direct feedback' in humans, it appears.  DeepMind's version of reinforcement learning that uses 'temporal value transport' to send a signal from reward backward, to shape actions, does better than alternative forms of neural networks.

It uses those signals to shape actions at the beginning of the funnel, a kind of feedback loop.  Also: Google's StarCraft II victory shows AI improves via diversity, invention, not reflexes They made a game of it, in other words.  They take simulated worlds, maps of rooms like you see in video games such as Quake and Doom, the kind of simulated environment that has become familiar in training of artificial agents.

The NMT was a way to make a computer search memory registers based not on explicit instructions but based simply on gradient descent in a deep learning network — in other words, learning the function by which to store and retrieve specific data.  The authors, Hung and colleagues, now take the approach of the NMT and, in a sense, bolt it onto normal RL.

Why StarCraft is the Perfect Battle Ground for Testing Artificial Intelligence

DeepMind, an offshoot of Google's parent company, debuted a computer program in January capable of beating professional players at one of the world's toughest video games.

StarCraft is a military science fiction franchise set in a universe rife with conflict, where armies of opponents face off to become the most powerful.

On its face, the video game has the standard hallmarks of its fantasy counterparts: strife in a post-apocalyptic world, a race to make yourself the most powerful opponent and a battle to defeat your enemies.

But instead of controlling a single first-person shooter agent, as in games like Halo or Overwatch, players manage a whole economy of builders, fighters and defense systems that work symbiotically to keep them from losing.

In 1997, IBM’s Deep Blue defeated the world chess champion, and other powerful computer algorithms, like DeepMind’s AlphaZero and AlphaGo, followed suit in defeating human board game masters at their craft.

On the other hand, Zerg spawns the quickest, but aren't strong fighters, so their power comes in numbers.  And besides simply selecting the strengths and weaknesses of your race, you also control multiple facets: workers gathering resources, builders creating defense systems, and fighters attacking enemies.

In 2011, Memorial University of Newfoundland computer scientist David Churchill co-authored a paper on build order in StarCraft II, studying how the prioritization of resource-building could affect success in the game.

The program he built for his doctoral thesis was designed to win a game called Ataxx, a 1990s-era arcade-style strategy game played on a virtual board.

It also makes it more difficult to see what your enemy is plotting — or, as Churchill says, engulfs you in the “fog of war.” “You don’t know what your enemy is doing until you’re standing right next to them,” he says.

And games like checkers or chess don’t happen in real time — once a player makes a move, there’s no time limit for an opponent to make theirs.

Churchill says video games can be a gateway to teaching machines to be better at image recognition, search suggestions, or any algorithm that has to assist humans in making decisions.

Since 2011, Churchill has organized a yearly, international event called the AIIDE StarCraft AI Competition, where game enthusiasts and professionals alike come together to build and test algorithms for games.

While the DeepMind team touts the algorithm’s high success rate, the amount of resources put into the project reaches a standard of power that goes well beyond the abilities of the average coder.

for example, the computer could see all its visible units without having to pan around the map to execute commands, and completed actions more precisely than a pro player clicking a mouse.

DeepMind created multiple automated players to specialize as certain races, and trained each by having them watch human game replays for 44 days.

After the Christmas break: Jog your mind with ML!

From: 2020-01-07 13:00 to: 15:00 Place: E:A, E-building, Ole Römers väg 3, LTH, Lund University - see map for how to enter the building.

Marc Deisenroth (University College London) and Shakir Mohamed (DeepMind) give a guest lecture each on their work within the area of Machine Learning

Our guests, Marc Deisenroth (University College London) and Shakir Mohamed (DeepMind) will give some insights into their insights in Machine Learning - and hopefully wake you up after the Holidays!  When: 7January 2020 13.00-15.00 Where: E:A, E-building, Ole Römers väg 3, LTH, Lund University Abstract: On our path toward fully autonomous systems, i.e., systems that operate in the real world without significant human intervention, reinforcement learning (RL) is a promising framework for learning to solve problems by trial and error.

While RL has had many successes recently, a practical challenge we face is its data inefficiency: In real-world problems (e.g., robotics) it is not always possible to conduct millions of experiments, e.g., due to time or hardware constraints.

Go master quits because AI 'cannot be defeated'

A master player of the Chinese strategy game Go has decided to retire, due to the rise of artificial intelligence that 'cannot be defeated'.

The 36-year-old former world champion started playing at the age of five, and turned pro just seven years later.

'On behalf of the whole AlphaGo team at DeepMind, I'd like to congratulate Lee Se-dol for his legendary decade at the top of the game, and wish him the very best for the future,' said Demis Hassabis, chief executive and co-founder of Deepmind.

player typically has a choice of 200 moves, compared with about 20 in chess - and there are more possible positions in Go than atoms in the universe, according to researchers.

Deepmind hopes that the development of AlphaGo will lead to 'similar techniques' that can be 'applied to other structured problems, such as protein folding, reducing energy consumption or searching for revolutionary new materials'.

Google's Deep Mind Explained! - Self Learning A.I.

Subscribe here: Become a Patreon!: Visual animal AI: .

Elon Musk on Google DeepMind | ARTIFICIAL INTELLIGENCE !!

ElonMusk has always warned us about #ArtificialIntelligence (AI) and this is what he thinks about Google's #DeepMind and #AI. . All the clips in the video are ...

DeepMind StarCraft II Demonstration

Join Artosis, RottterdaM and a cast of special guests for a unique StarCraft II showcase live from DeepMind in London, in partnership with Blizzard.

DeepMind’s AlphaStar: A Grandmaster Level StarCraft 2 AI

Check out Weights & Biases here and sign up for a free demo: Their mentioned blog post: ..

How Google's DeepMind is Using AI to Tackle Climate Change

AI is often seen as a magical fix-all. It's not, but it is, says Sims Witherspoon, a powerful tool for unlocking communities' problem-solving capacities. Witherspoon ...

DeepMind's AI Takes An IQ Test

The paper "Measuring abstract reasoning in neural networks" is available here: Pick up cool perks ..

Google's DeepMind AI Just Taught Itself To Walk

Google's artificial intelligence company, DeepMind, has developed an AI that has managed to learn how to walk, run, jump, and climb without any prior guidance ...

Deepmind artificial intelligence @ FDOT14

Smartphone bad audio quality Deepmind technology wrecking old Atari games, presented by neuroscientist and game developper Demis Hassabis, Deepmind ...

DeepMind - The Role of Multi-Agent Learning in Artificial Intelligence Research

Thore Graepel is a Research Scientist at Google DeepMind, and Professor of Computer Science at UCL. Recorded: March, 2017.