Tag Archives: Mike
#437269 DeepMind’s Newest AI Programs Itself ...
When Deep Blue defeated world chess champion Garry Kasparov in 1997, it may have seemed artificial intelligence had finally arrived. A computer had just taken down one of the top chess players of all time. But it wasn’t to be.
Though Deep Blue was meticulously programmed top-to-bottom to play chess, the approach was too labor-intensive, too dependent on clear rules and bounded possibilities to succeed at more complex games, let alone in the real world. The next revolution would take a decade and a half, when vastly more computing power and data revived machine learning, an old idea in artificial intelligence just waiting for the world to catch up.
Today, machine learning dominates, mostly by way of a family of algorithms called deep learning, while symbolic AI, the dominant approach in Deep Blue’s day, has faded into the background.
Key to deep learning’s success is the fact the algorithms basically write themselves. Given some high-level programming and a dataset, they learn from experience. No engineer anticipates every possibility in code. The algorithms just figure it.
Now, Alphabet’s DeepMind is taking this automation further by developing deep learning algorithms that can handle programming tasks which have been, to date, the sole domain of the world’s top computer scientists (and take them years to write).
In a paper recently published on the pre-print server arXiv, a database for research papers that haven’t been peer reviewed yet, the DeepMind team described a new deep reinforcement learning algorithm that was able to discover its own value function—a critical programming rule in deep reinforcement learning—from scratch.
Surprisingly, the algorithm was also effective beyond the simple environments it trained in, going on to play Atari games—a different, more complicated task—at a level that was, at times, competitive with human-designed algorithms and achieving superhuman levels of play in 14 games.
DeepMind says the approach could accelerate the development of reinforcement learning algorithms and even lead to a shift in focus, where instead of spending years writing the algorithms themselves, researchers work to perfect the environments in which they train.
Pavlov’s Digital Dog
First, a little background.
Three main deep learning approaches are supervised, unsupervised, and reinforcement learning.
The first two consume huge amounts of data (like images or articles), look for patterns in the data, and use those patterns to inform actions (like identifying an image of a cat). To us, this is a pretty alien way to learn about the world. Not only would it be mind-numbingly dull to review millions of cat images, it’d take us years or more to do what these programs do in hours or days. And of course, we can learn what a cat looks like from just a few examples. So why bother?
While supervised and unsupervised deep learning emphasize the machine in machine learning, reinforcement learning is a bit more biological. It actually is the way we learn. Confronted with several possible actions, we predict which will be most rewarding based on experience—weighing the pleasure of eating a chocolate chip cookie against avoiding a cavity and trip to the dentist.
In deep reinforcement learning, algorithms go through a similar process as they take action. In the Atari game Breakout, for instance, a player guides a paddle to bounce a ball at a ceiling of bricks, trying to break as many as possible. When playing Breakout, should an algorithm move the paddle left or right? To decide, it runs a projection—this is the value function—of which direction will maximize the total points, or rewards, it can earn.
Move by move, game by game, an algorithm combines experience and value function to learn which actions bring greater rewards and improves its play, until eventually, it becomes an uncanny Breakout player.
Learning to Learn (Very Meta)
So, a key to deep reinforcement learning is developing a good value function. And that’s difficult. According to the DeepMind team, it takes years of manual research to write the rules guiding algorithmic actions—which is why automating the process is so alluring. Their new Learned Policy Gradient (LPG) algorithm makes solid progress in that direction.
LPG trained in a number of toy environments. Most of these were “gridworlds”—literally two-dimensional grids with objects in some squares. The AI moves square to square and earns points or punishments as it encounters objects. The grids vary in size, and the distribution of objects is either set or random. The training environments offer opportunities to learn fundamental lessons for reinforcement learning algorithms.
Only in LPG’s case, it had no value function to guide that learning.
Instead, LPG has what DeepMind calls a “meta-learner.” You might think of this as an algorithm within an algorithm that, by interacting with its environment, discovers both “what to predict,” thereby forming its version of a value function, and “how to learn from it,” applying its newly discovered value function to each decision it makes in the future.
Prior work in the area has had some success, but according to DeepMind, LPG is the first algorithm to discover reinforcement learning rules from scratch and to generalize beyond training. The latter was particularly surprising because Atari games are so different from the simple worlds LPG trained in—that is, it had never seen anything like an Atari game.
Time to Hand Over the Reins? Not Just Yet
LPG is still behind advanced human-designed algorithms, the researchers said. But it outperformed a human-designed benchmark in training and even some Atari games, which suggests it isn’t strictly worse, just that it specializes in some environments.
This is where there’s room for improvement and more research.
The more environments LPG saw, the more it could successfully generalize. Intriguingly, the researchers speculate that with enough well-designed training environments, the approach might yield a general-purpose reinforcement learning algorithm.
At the least, though, they say further automation of algorithm discovery—that is, algorithms learning to learn—will accelerate the field. In the near term, it can help researchers more quickly develop hand-designed algorithms. Further out, as self-discovered algorithms like LPG improve, engineers may shift from manually developing the algorithms themselves to building the environments where they learn.
Deep learning long ago left Deep Blue in the dust at games. Perhaps algorithms learning to learn will be a winning strategy in the real world too.
Image credit: Mike Szczepanski / Unsplash Continue reading
Posted in Human Robots
#435174 Revolt on the Horizon? How Young People ...
As digital technologies facilitate the growth of both new and incumbent organizations, we have started to see the darker sides of the digital economy unravel. In recent years, many unethical business practices have been exposed, including the capture and use of consumers’ data, anticompetitive activities, and covert social experiments.
But what do young people who grew up with the internet think about this development? Our research with 400 digital natives—19- to 24-year-olds—shows that this generation, dubbed “GenTech,” may be the one to turn the digital revolution on its head. Our findings point to a frustration and disillusionment with the way organizations have accumulated real-time information about consumers without their knowledge and often without their explicit consent.
Many from GenTech now understand that their online lives are of commercial value to an array of organizations that use this insight for the targeting and personalization of products, services, and experiences.
This era of accumulation and commercialization of user data through real-time monitoring has been coined “surveillance capitalism” and signifies a new economic system.
Artificial Intelligence
A central pillar of the modern digital economy is our interaction with artificial intelligence (AI) and machine learning algorithms. We found that 47 percent of GenTech do not want AI technology to monitor their lifestyle, purchases, and financial situation in order to recommend them particular things to buy.
In fact, only 29 percent see this as a positive intervention. Instead, they wish to maintain a sense of autonomy in their decision making and have the opportunity to freely explore new products, services, and experiences.
As individuals living in the digital age, we constantly negotiate with technology to let go of or retain control. This pendulum-like effect reflects the ongoing battle between humans and technology.
My Life, My Data?
Our research also reveals that 54 percent of GenTech are very concerned about the access organizations have to their data, while only 19 percent were not worried. Despite the EU General Data Protection Regulation being introduced in May 2018, this is still a major concern, grounded in a belief that too much of their data is in the possession of a small group of global companies, including Google, Amazon, and Facebook. Some 70 percent felt this way.
In recent weeks, both Facebook and Google have vowed to make privacy a top priority in the way they interact with users. Both companies have faced public outcry for their lack of openness and transparency when it comes to how they collect and store user data. It wasn’t long ago that a hidden microphone was found in one of Google’s home alarm products.
Google now plans to offer auto-deletion of users’ location history data, browsing, and app activity as well as extend its “incognito mode” to Google Maps and search. This will enable users to turn off tracking.
At Facebook, CEO Mark Zuckerberg is keen to reposition the platform as a “privacy focused communications platform” built on principles such as private interactions, encryption, safety, interoperability (communications across Facebook-owned apps and platforms), and secure data storage. This will be a tough turnaround for the company that is fundamentally dependent on turning user data into opportunities for highly individualized advertising.
Privacy and transparency are critically important themes for organizations today, both for those that have “grown up” online as well as the incumbents. While GenTech want organizations to be more transparent and responsible, 64 percent also believe that they cannot do much to keep their data private. Being tracked and monitored online by organizations is seen as part and parcel of being a digital consumer.
Despite these views, there is a growing revolt simmering under the surface. GenTech want to take ownership of their own data. They see this as a valuable commodity, which they should be given the opportunity to trade with organizations. Some 50 percent would willingly share their data with companies if they got something in return, for example a financial incentive.
Rewiring the Power Shift
GenTech are looking to enter into a transactional relationship with organizations. This reflects a significant change in attitudes from perceiving the free access to digital platforms as the “product” in itself (in exchange for user data), to now wishing to use that data to trade for explicit benefits.
This has created an opportunity for companies that seek to empower consumers and give them back control of their data. Several companies now offer consumers the opportunity to sell the data they are comfortable sharing or take part in research that they get paid for. More and more companies are joining this space, including People.io, Killi, and Ocean Protocol.
Sir Tim Berners Lee, the creator of the world wide web, has also been working on a way to shift the power from organizations and institutions back to citizens and consumers. The platform, Solid, offers users the opportunity to be in charge of where they store their data and who can access it. It is a form of re-decentralization.
The Solid POD (Personal Online Data storage) is a secure place on a hosted server or the individual’s own server. Users can grant apps access to their POD as a person’s data is stored centrally and not by an app developer or on an organization’s server. We see this as potentially being a way to let people take back control from technology and other companies.
GenTech have woken up to a reality where a life lived “plugged in” has significant consequences for their individual privacy and are starting to push back, questioning those organizations that have shown limited concern and continue to exercise exploitative practices.
It’s no wonder that we see these signs of revolt. GenTech is the generation with the most to lose. They face a life ahead intertwined with digital technology as part of their personal and private lives. With continued pressure on organizations to become more transparent, the time is now for young people to make their move.
Dr Mike Cooray, Professor of Practice, Hult International Business School and Dr Rikke Duus, Research Associate and Senior Teaching Fellow, UCL
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Image Credit: Ser Borakovskyy / Shutterstock.com Continue reading
Posted in Human Robots