Daily Archives: March 2, 2017

In an era of maturing artificial intelligence technology, what does the future of the corporation look like? Will the rise of robots help us do our jobs better, or harm them ? This dynamic has become a mainstay of the dialogue around AI, with voices from technology visionaries such as Bill Gates and Stephen Hawking weighing in.

But at Fortunes Most Powerful Women International Summit in Hong Kong on Tuesday, leaders at two of the worlds most powerful tech giants pushed back on those concerns. AI is intended to helpnot hinderthe human workforce, they said.

AI is actually not new for us, said Vanitha Narayanan, chairman of IBM India, whose Watson supercomputer has risen to global acclaim. But technology always comes way ahead of policy.

Just as her boss, IBM CEO Ginni Rometty, has insisted "it will not be man or machine," Narayanan said that many companies, especially service-oriented firms, benefit from the technology.

Consider a bank in Brazil that has 50 different products, she said. The average call center rep cannot handle 50 products. So theyre using Watson in their call centernot to replace the call center rep, but actually help the call center rep understand their products and their offerings and be able to serve that customer much better than he or she may have been able to.

Leonie Valentine, managing director of sales and operations at Google Hong Kong, likened use of artificial intelligence to a murmuration, when a starling flock take flight without any kind of preprogrammed system.

The birds just know how to find how to find formation in a murmuration. And it just kind of works, Valentine said. Its a fantastic analogy for what the future of organization will look like.

Subscribe to the Broadsheet, Fortunes newsletter for powerful women.

The Google executive also defended artificial intelligence as a means of supplementing human workers.

Take the example of call center operators, When you go and survey your customers about where they would like your staff to spend your time," she said, "its not on problem-solving, troubleshooting, credit issues, or billing, right? Its Help me, advise me, send me on the trip that I wanted to go to.

Valentine said that since a lot of the technology has been put in place, well actually finally get to the place thats been Nirvana for the last 20 or 30 years in corporations, which is moving people to high-value tasks, which is actually taking the intelligence of the organization and forming that human intelligence, and where we need value-based judgments and real-time decision-making and a human touch, putting that back into the hands of the customer.

Said Valentine: Thats the sort of stuff Im looking forward to."

View original post here:

Google and IBM: We Want Artificial Intelligence to Help You, Not Replace You - Fortune

Gone are the days of the hobbyist hackermodern malware is a trillion-euro business.

Dr Eva Maia at VisionTechLab, a young cybersecurity firm in Matosinhos, Portugal, said that attacks on computer networks are not only multiplying, they are also growing sneakier.

Malwares typically go unnoticed for months by remaining dormant on infected computers, said Dr Maia. This was recently the case in the Panama Papers attack, where no one knew that the network had been compromised until long after the damage was done.

In the EU-funded SecTrap project, VisionTechLab has been studying the market for a new line of defence that could rob malicious software of its current hiding places.

Conventional antiviruses and firewalls are trained like nightclub bouncers to block known suspects from entering the system. But new threats can be added to the wanted list only after causing trouble. If computers could instead be trained as detectives, snooping around their own circuits and identifying suspicious behaviour, hackers would have a harder time camouflaging their attacks.

The challenge is that machines have traditionally been built to follow orders, not recognise patterns or draw conclusions. Dr Maia is working on advances in artificial intelligence (AI) to change that.

Bootcamp

We are seeing a boom in AI techniques, said Dr Maia. Research that was previously theoretical is now moving from academic laboratories to industry at an unprecedented pace.

Research that was previously theoretical is now moving from academic laboratories to industry at an unprecedented pace.

Dr Eva Maia, VisionTechLab, Portugal

Over the past few years, computers have started driving passenger cars,following voice orders and outmatching humans at identifying faces on photographs. These breakthroughs are the fruit of a new trend in AI based on mimicking living neural networks.

In the same way as our brains sort new information based on past experiences, enough practice data can teach computers to learn, categorise and generalise for themselves.

How many examples are needed to identify a trend can run into astronomical numbers. Fortunately, vast hoards of behavioural data are strewn everyday across the internet by heedless bloggers, commentators and social media users.

Computers learnt their first cognitive functions by devouring terabytesof this online text, sound and images. With the help of recent computing power and the kind of algorithms developed by Dr Maia, they have become so good at identifying content that they now label some of it for us.

The challenge in cybersecurity is to use this ability to distinguish between innocent and malicious behaviour on a computer. For this, Alberto Pelliccione, chief executive of ReaQta, a cybersecurity venture in Valletta, Malta, has found an analogous way of educating by experience.

ReaQta breeds millions of malware programs in a virtual testing environment known as a sandbox, so that algorithms can inspect their antics at leisure and in safety. It is not always necessary to know what they are trying to steal. Just to record the applications that they open and their patterns of operation can be enough.

So that the algorithms can learn about business as usual, they then monitor the behaviour of legal software, healthy computers, and ultimately the servers of each new client. Their lesson never ends. The algorithms continue to learn from their users even after being put into operation.

In doing so, ReaQtas algorithms can assess whether programs or computers are behaving unusually. If they are, they inform human operators, who can either shut them down or study the tactics of the malware infecting them. The objective of the artificial intelligence is not to teach computers what we define as good or bad data, but to spot anomalies, said Pelliccione.

Nowhere to hide

This is welcome news for IT administrators. Cyber criminals typically attack the computer networks of large organisations by compromising the machines of less security-savvy users on their periphery and working their way through to the centre. A few weak links in a sprawling network are difficult to spot and can progressively put an entire company at risk.

To make matters worse, hackers install dormant access points on each machine that they compromise. If security analysts manage to block one, hackers return through another. Dormant access points are notoriously difficult to spot because they do nothing until hackers activate them.

As part of the European ProBOS project, ReaQta has developed software that can be nested at the very core of machines, between their operating system and hardware. Its role is to monitor daily operations in every corner of the system, allowing AI algorithms to sift through ubiquitous data and spot any malicious installation.

ReaQta is licensing the security platform across European and Southeast Asian markets this month. Its first clients are companies that operate over 500 computers simultaneously.

Next year, VisionTechLab plans to release its first AI security services for banks and governments. In the longer term, Dr Maia sees applications for individuals.

For all the benefits of mobile devices, social networking and cloud computing, these technologies are placing more private data at risk. While AI may not yet be capable of guaranteeing its safety, it can now shine a powerful search light on any attempts to steal it.

If you liked this article, please consider sharing it on social media.

Cybersecurity is at the heart of the EU's strategy for the Digital Single Market.

The EU's cybersecurity strategy was created to embed cybersecurity into new policies in areas such as automated driving, make the EU a strong player in the cybersecurity market, and ensure that all Member States have similar capabilities to fight cyber crime.

View post:

Artificial intelligence being turned against spyware - Horizon magazine

Facebook has started using artificial intelligence to identify users who are potentially at risk of taking their own lives.

The social network has developed algorithms capable of scanning posts and comments for warning signs.

These could be phrases such as Are you okay? or Im worried about you, or more general talk of sadness and pain.

The AI tool would send such posts to a human review team, which would get in touch with the user thought to be at risk and offer help, in the form of contact details for support services or a chat with a member of staff through Facebook Messenger.

The site had previously relied on other users reporting worrying updates.

The AI is actually more accurate than the reports that we get from people that are flagged as suicide and self injury, Facebook product manager Vanessa Callison-Burch told BuzzFeed. The people who have posted that content [that AI reports] are more likely to be sent resources of support versus people reporting to us.

The system is currently being tested in the US.

The site has also announced new safety features for Facebook Live, which has been used to live stream several suicides.

Users can now flag up concerning Facebook Live behaviour with the site, which will display advice and highlight the video to staff for immediate review.

The goal is to provide help as quickly as possible, mid-broadcast rather than post-broadcast.

Some might say we should cut off the stream of the video the moment there is a hint of somebody talking about suicide, said Jennifer Guadagno, the projects lead researcher.

But what the experts emphasised was that cutting off the stream too early would remove the opportunity for people to reach out and offer support. So, this opens up the ability for friends and family to reach out to a person in distress at the time they may really need it the most.

Facebook CEO Mark Zuckerberg described plans to use AI to identify worrying content in a recently published manifesto.

Looking ahead, one of our greatest opportunities to keep people safe is building artificial intelligence to understand more quickly and accurately what is happening across our community, it read.

An earlier version of the piece said that it would take many years to develop AI systems capable of identifying issues such as bullying and terrorism risks online, but the section was removed before the manifesto was publicly issued.

See the original post:

Facebook using artificial intelligence to help suicidal users - The Independent

You have to know when to hold em and when to fold em.

Now a program developed by computer scientist at the University of Alberta can do both and do it much better than an entire cohort of professional poker players.

The achievement marks a new milestone for artificial intelligence (AI) involving deep learning, a style of programming that mimics certain aspects of how human brains acquire expertise. But while the program, dubbed DeepStack, represents a significant step, a rival U.S. team said that the method by which it was tested against humans was insufficient to reveal the true extent of its capabilities.

Games provide an important testbed for artificial intelligence because they offer a well-defined arena where programming approaches can be evaluated and compared. Last year, another deep learning system developed by Google DeepMind managed to beat the world champion at Go, a board game that is fiendishly complex despite its simple rules because of the number of possible decisions a player can make.

Poker specifically the version known as Heads-Up No-Limit Texas Holdem presents a different kind of challenge. Unlike Go or chess, where both players can assess the state of the game simply by looking at the board, poker players must deal with incomplete knowledge because of cards that are hidden from view.

The essence of poker is being able to make decisions when you dont have all of the information that you need, said Michael Bowling, who leads the universitys computer poker research group. Dr. Bowling added that the same kind of reasoning is often required when computers have to solve real-world problems, which makes poker an attractive hurdle for designers of intelligent systems.

The Alberta group has been working for 20 years on programs that try to solve poker. In 2008, it developed an algorithm that could defeat top human players at the heads-up limit version of the game, in which all bets are of fixed size. There are one thousand billion different decision points than can arise in such a game, a numerical challenge that Dr. Bowling compares to checkers. While not trivial, its a game that a computer can be hardwired to win.

In comparison, the no-limit version of the game is astronomically more complicated because players can choose to bet any amount up to the number of chips in their possession. A winning strategy often involves betting high when the opponent believes incorrectly that his or her hand is the stronger one.

In designing DeepStack, Dr. Bowlings team, together with colleagues at the Czech Technical University in Prague, had to create a system that not only understood the strength of its own hand and make an informed guess about its opponents, but also weigh what its opponent might be thinking in order to bluff and conceal its own intentions.

People think of bluffing as this very human, psychological thing, but it pretty much falls out of the mathematics of the game, Dr. Bowling said. He added that DeepStack had to be able to learn how to bluff, otherwise it would be a terrible player.

The team developed a deep-learning system that tried to make the best choice by looking only a few actions ahead, otherwise it would be overwhelmed by the mathematical possibilities. The system was trained using an army of lesser computers who played through a multitude of game scenarios, gradually building up DeepStacks intuition for what to do. An overview of how the program works along with the results of its matchups against human players were published Thursday in the journal Science.

To test the system, the team recruited 33 professional poker players from 17 countries to go toe-to-toe with DeepStack, with an offer of cash prizes up to $5,000 awarded to the top three players. The players were each given four weeks in late 2016 to complete 3,000 games against the program. Only a third of the human players went the full distance. Of those, all but one were beaten by a significant enough margin to rule out luck.

Tuomas Sandholm, who leads the Alberta groups chief competitor at Carnegie Mellon University in Pittsburgh, said that DeepStack featured a new combination of programming methods that made it a potent player.

However, he cited several weaknesses in the way the system was tested, including the fact that the players DeepStack faced were not the worlds best and the prizes they were offered were likely not sufficient to motivate the players to perform at their sharpest. He also said that 3,000 matches would not provide humans with enough experience to learn to adjust and potentially outsmart the program.

Dr. Sandholms team has been working with a different system called Libratus that runs on a supercomputer and does not employ deep learning. but instead uses a trial-and-error approach called reinforcement learning. In a sign of how close the competition has become, last month Libratus beat a team of four top, human Heads-Up No-Limit Texas Holdem players. There are no plans as yet for a tournament that would pit the two systems against each other.

In the meantime, Dr. Bowling said there was plenty of scope to beef up DeepStacks capabilities and also to try it out on different variations of no-limit poker that more closely resemble a human championship game.

Follow Ivan Semeniuk on Twitter: @ivansemeniuk

Read the original:

Artificial intelligence crosses the poker barrier - The Globe and Mail