Daily Archives: May 18, 2017

Were currently in the final phase of AI development, in which weve advanced past teaching computers to follow rules and evaluating the best solutions to problems. With a clearly defined goal and enough data to parse, AI can now learn how to arrive at a solution on its own.

Thats both exciting and terrifying all at once, says data scientist and YouTuber Siraj Raval on stage at TNW Conference in Amsterdam. According to him, the danger we now face in the 21st century is that governments and corporations can arm themselves with powerful AI to control societies, and we need a way to tackle that.

What are we afraid of, specifically? Raval explains that there are two ways for AI to destroy us. The first is annihilation, Terminator-style but its less about robots being intrinsically evil and more about them seeing humans as the problem and opting to eliminate us.

The other is manipulation. If youve been following the issue of fake news in the media lately, you know where this is going. AI can be used to generate bogus information whether its a Wikipedia article or even a video that looks like a politician addressing the public thats indistinguishable from genuine sources of news and facts.

The way to prevent AI from killing us all, Raval says, is to democratize AI and put it in the hands of as many people as possible. That way, no single entity controls that superpower and has an unfair advantage over its enemies and subjects.

Organizations like Google, Elon Musks OpenAI, Microsoft and Good AI are already working on various initiatives to make this happen.

At a more basic level, Raval notes, people need to understand, at a high level, how AI works and what theyre comfortable with.

By doing so, well be able to have more conversations not just about the capabilities of AI, but also about how we want to harness its power, and how we want to limit it.

Read next: Europes fastest-growing young tech companies: announcing the European Tech30 2017

Read the original post:

How to keep AI from killing us all - TNW

Meet These Incredible Women Advancing AI Research Forbes This list of 20+ leading women in AI research is not comprehensive. Far more talented people contribute to the field than we can quickly summarize in a single article. All of the women featured here overcame personal and professional challenges to ...

Read more here:

Meet These Incredible Women Advancing AI Research - Forbes

How can humans stay ahead of an ever growing machine intelligence? I think the challenge for usis to always be creative, saysformer world chess champion Garry Kasparov, speaking on stage here at TechCrunch Disrupt New York.

Its twenty years since Kasparov lost to Deep Blue, IBMs supercomputer. The 1997 AI stress test equivalent of DeepMinds AlphaGo beating Lee Sedol last year.

I would remind people that I won the first match, quipped Kasparov when the historic defeat was brought up.

There are certain things that we can do that machines are not at all going to be able to replicate, he said. Its things like love, emotions, passions One of the rules is anything that we know how we do machines can do it. But theres so many things that we do that we dont know how we do and thats where we humans can excel.

Machines dont have understanding, machines dont recognize strategic patterns, and machines dont have a purpose, he added, arguing for a blend of what humans and machines do best andfinding a way to bring together these complementary skills.

The new generation of self-learning AIs being built by the likes ofDeepMind alsopresents another challengefor humans, in that it is much harder to understand thedecisions of these algorithms, said Kasparov, flagging this as one of the biggest challenges for applying this type of tech.

You cannot identify why iteration five is better than iteration 10, he said. With Deep Blue if you have a couple of years to spare you can find out why a certain move was there. With AlphaGo the problem is you can only guess. And I think thats one of the biggest challenges with the new generation of AI.

Whilehe wasgenerally upbeat about the combined possibilities of humans and machines touching on his Advanced Chessconcept, for example, in which machines and humans play together as a team he also warned over some of thesocietal threats here.

Technology isa double-edged weapon, he said, pointing to regimes like Putins in Russia appropriating techtools to try to undermine the very foundations of the free world and to topple dissent in their own countries.

We should realize the dangers of Internet and free space being polluted by fake news and the misinformation run by the [rogue] states mainly of course Putins regime that is so sophisticated, he said, adding: Putin doesnt care who he supports outside Russiato spread chaos and to create more uncertainty and to undermine democracyas an institution and even as a concept.

He also discussed thethreat that increasingly capable AI poses to (human) jobs, arguing that we should not try to predict what will happen in the future but rather look atimmediate problems.

We make predictions and most arewrong because were trying to base iton our past experience, he argued. I think the problem is not that AI is exceeding our level. Its another cycle. Machines have been constantly replacing all sorts of jobs We should not talk about jobs being lost but people suffering.

We have to look for new opportunities. In my opinion the biggest challenge is not that the jobs are being lost, but the challenge is that they are not being lost fast enough. Because unless you have a cycle moving fast you will not be able to create new sustainable jobs. You will haveto create a new growth that will help toreplace jobs that are being lost.

AI is just another challenge. The difference is that now intelligence machines are coming after people with a college degree or with social media andTwitter accounts, he added.

Read more:

We should not talk about jobs being lost but people suffering, says Kasparov on AI - TechCrunch

When people work together on a project, they often come to think theyve figured out the problems in their own respective spheres. If trouble persists, its somebody elseengineering, say, or the marketing departmentwho is screwing up. That local focus means finding the best way forward for the overall project is often a struggle. But what if adding artificial intelligence to the conversation, in the form of a computer program called a bot, could actually make people in groups more productive?

This is the tantalizing implication of a study published Wednesday in Nature. Hirokazu Shirado and Nicholas Christakis, researchers at Yale Universitys Institute for Network Science, were wondering what would happen if they looked at artificial intelligence (AI) not in the usual wayas a potential replacement for peoplebut instead as a useful companion and helper, particularly for altering human social behavior in groups.

First the researchers asked paid volunteers arranged in online networks, each occupying one of 20 connected positions, or nodes, to solve a simple problem: Choose one of three colors (green, orange or purple) with the individual, or local, goal of having a different color from immediate neighbors, and the collective goal of ensuring that every node in the network was a different color from all of its neighbors. Subjects pay improved if they solved the problem quickly. Two thirds of the groups reached a solution in the allotted five minutes and the average time to a solution was just under four minutes. But a third of the groups were still stymied at the deadline.

The researchers then put a botbasically a computer program that can execute simple commandsin three of the 20 nodes in each network. When the bots were programmed to act like humans and focused logically on resolving conflicts with their immediate neighbors, they didnt make much difference. But when the researchers gave the bots just enough AI to behave in a slightly noisy fashion, randomly choosing a color regardless of neighboring choices, the groups they were in solved the problem 85 percent of the timeand in 1.7 minutes on average, 55.6 percent faster than humans alone.

Being just noisy enoughmaking random color choices about 10 percent of the timemade all the difference, the study suggests. When a bot got much noisier than that, the benefit soon vanished. A bots influence also varied depending on whether it was positioned at the center of a network with lots of neighbors or on the periphery.

So why would making what looks like the wrong choicein other words, a mistakeimprove a groups performance? The immediate result, predictably, was short-term conflict, with the bots neighbors in effect muttering, Why are you suddenly disagreeing with me? But that conflict served to nudge neighboring humans to change their behavior in ways that appear to have further facilitated a global solution, the co-authors wrote. The humans began to play the game differently.

Errors, it seems, do not entirely deserve their bad reputation. There are many, many natural processes where noise is paradoxically beneficial, Christakis says. The best example is mutation. If you had a species in which every individual was perfectly adapted to its environment, then when the environment changed, it would die. Instead, random mutations can help a species sidestep extinction.

Were beginning to find that errorand noisy individuals that we would previously assume add nothingactually improve collective decision-making, says Iain Couzin, who studies group behavior in humans and other species at the Max Planck Institute for Ornithology and was not involved in the new work. He praises the deliberately simplified model used in the Nature study for enabling the co-authors to study group decision-making in great detail, because they have control over the connectivity. The resulting ability to minutely track how humans and algorithms collectively make decisions, Couzin says, is really going to be the future of quantitative social science.

But how realistic is it to think human groups will want to collaborate with algorithms or botsespecially slightly noisy onesin making decisions? Shirado and Christakis informed some of their test groups that they would be partnering with bots. Perhaps surprisingly, it made no difference. The attitude was, I don't care that youre a bot if youre helping me do my job, Christakis says. Many people are already accustomed to talking with a computer when they call an airline or a bank, he adds, and the machine often does a pretty good job. Such collaborations are almost certain to become more common amid the increasing integration of the internet with physical devices, from automobiles to coffee makers.

Real-world, bot-assisted company meetings might not be too far behind. Business conferences already tout blended digital and in-person events, featuring what one conference planner describes as integrated online and offline catalysts that use virtual reality, augmented reality and artificial intelligence. Shirado and Christakis suggest slightly noisy bots are also likely to turn up in crowdsourcing applicationsfor instance, to speed up citizen science assessment of archaeological or astronomical images. They say such bots could also be useful in social mediato discourage racist remarks, for example.

But last year when Microsoft introduced a twitter bot with simple AI, other users quickly turned it into epithetspouting bigot. And the opposite concern is that mixing humans and machines to improve group decision-making could enable businessesor botsto manipulate people. Ive thought a lot about this, Christakis says. You can invent a gun to hunt for food or to kill people. You can develop nuclear energy to generate electric power or make the atomic bomb. All scientific advances have this Janus-like potential for evil or good.

The important thing is to understand the behavior involved, so we can use it to good ends and also be aware of the potential for manipulation, Couzin says. Hopefully this new research will encourage other researchers to pick up on this idea and apply it to their own scenarios. I dont think it can be just thrown out there and used willy-nilly.

Original post:

Pushy AI Bots Nudge Humans to Change Behavior - Scientific ... - Scientific American

Forbes

AI Is Taking Away Jobs, But The Rise Of More Multibillion-Dollar Companies Is Just Starting Forbes For better or worse, tech innovation is on course to replace us all one day. Economists have been grappling with this since machinery started replacing human labor during the Industrial Revolution. But recently, the chatter seems to have taken on a ...

Read more here:

AI Is Taking Away Jobs, But The Rise Of More Multibillion-Dollar Companies Is Just Starting - Forbes

At one time or another, most lawyers involved in eDiscovery have felt the unique pressure induced by a slow-moving document review. That pressure to speed a review is one of the reasons that eDiscovery is effectively ground zero for todays exploding use of AI in law.

Many lawyers first encounter with machine-learning technology was through technology-assisted review (TAR) or predictive coding. Those technologies marked the first significant use of AI in any field of law, and that use is expanding.

So How Does TAR Lighten the Load?

Consider this scenario: Youre halfway through a document review and running out of time. Would it help to switch over to TAR?

Answer: Depending on the size of the collection, yes. In fact, because youve already coded half the documents, you can use those coding decisions to jumpstart the TAR process.

A law firm recently faced virtually the identical scenario. The firm had to review a relatively small collection of about 40,800 documents in a short time. As it waited for approval from its client to use TAR, the firm dove into the review, knowing time was tight. By the time the client gave the green light to use TAR, the firm had coded 18,200 of the 40,800 documents.

The documents theyd coded provided a ready-made set of seeds to use to train the TAR algorithm. This is possible with a TAR engine that uses continuous active learning, a machine learning protocol that enables it to use any and all previously coded documents as judgmental seeds to start the process.

The coded documents were fed into the system and then, based on that input, the entire population was analyzed and ranked. From that point, the TAR engine started feeding the reviewers batches of 50 documents each. Each batch contained the next-best documents that were most likely to be responsive. In each batch, the tool also included a few contextually diverse documents to make sure there are no topics or concepts left unexplored.

As the reviewers completed their batches, the system continuously used their judgments to re-rank the entire population, improving its rankings with each new batch. The review proceeded along this track until the reviewers started seeing batches with few, if any, relevant documents, which told them that few relevant documents remained. Testing showed that the review had achieved high recall, meaning the reviewers had found the vast majority of the relevant documents.

In the end, the law firm had reviewed another 6,800 documents beyond the 18,000 theyd reviewed before starting TAR. That meant there were another 15,800 documents they did not have to review. So, from the time they started using TAR, the process cut out 70% of the work that manual review would have required. The firm calculated that this saved its clients more than $70,000.

This scenario of how TAR works focuses on just one of 20 questions answered in a new book, Ask Catalyst: A Users Guide to TAR, which is available for free download at this link.

Read more:

How AI And Machine-Learning Tools Lighten The eDiscovery Load - Above the Law

So, in partnership with Linkedin, Monster, CareerBuilder, Glassdoor, and -- surprisingly -- Facebook, the tech giant will be launching Google for Jobs, an AI-powered search engine that combines Google search, machine learning (to delve into career sites), job boards, staffing agencies and applicant tracking systems to help you find work in your area. You can even set an alert for your search. That means, if the barista position you're looking for isn't available today, Google will notify you when it surfaces.

That's outstanding if you're one the 4.4 percent of the population that's unemployed. It was also good to see Google CEO Sundar Pichai demo job search with retail jobs instead of developer jobs. There are plenty of ways for a computer science major to find a gig. Finding a spot as a cashier in a store can be a bit trickier because of inconsistent job titles and some companies sticking to the same job posting plan year after year, even though better solutions are out there. Google takes all the information from various sources, throws its AI at it, and spits out an easy to read list that's beneficial to everyone involved.

This announcement is also is a timely reminder of the importance of smartphones for those struggling to make a living. The ubiquitous device has become an important tool for finding work, as for many it's their primary means of getting online. With Google's announcement, that task is now just a bit easier.

Google's AI initiative will power some pretty cool technology like Google Lens and Assistant, but it's good to know the company is also making sure that artificial intelligence is helping folks that need a paycheck more than they need to know what kind of flower they're looking at.

For all the latest news and updates from Google I/O 2017, follow along here.

See the article here:

AI-powered Google for Jobs has work for everybody - Engadget