Monthly Archives: March 2017

On a chilly March afternoon last year in the South Korean capital Seoul, a computer algorithm made history.

A program called AlphaGo beat the reigning human world champion at go, an ancient Chinese board game considered to be one of the most complex pastimes man has ever devised.

The game has remained an inviolably human pursuit for centuries, and one of the hardest challenges for artificial intelligence (AI) because of the vast number of possible moves more than the number of atoms in the universe and the need to employ creativity to win.

In Seouls Four Seasons hotel, AlphaGos victory over five games was ruthless: Lee Sedol, the 33-year-old human go grandmaster, lost 4-1. At a press conference afterwards, he said with a trace of wonder: Today, I am speechless.

Just two months earlier, AlphaGo had been featured on the cover of Nature, the premier peer-reviewed scientific journal, having defeated the human European go champion 5-0. That and its triumph over Lee cemented its position as a rare scientific breakthrough that came years ahead of scientists predictions.

This is the first time that a computer program has defeated a human professional player in the full-sized game of go, a feat previously thought to be at least a decade away, the team behind it wrote.

AlphaGo is the brainchild of DeepMind Technologies, a London-based AI company acquired by Google in 2014 for 400m. The AlphaGo feature was the second time in a year DeepMind had made the cover of Nature. Ten months later, last October, the team made a third appearance in the journal, making them singularly prolific among their academic peers.

With its cadre of researchers, from Bayesian mathematicians to cognitive neuroscientists, statisticians and computer scientists, DeepMind has amassed arguably the most formidable community of world-leading academics specialising in machine intelligence anywhere in the world.

What we are trying to do is a unique cultural hybrid the focus and energy you get from start-ups with the kind of blue-sky thinking you get from academia, says Demis Hassabis, co-founder and chief executive. Weve hired 250 of the worlds best scientists, so obviously theyre here to let their creativity run riot, and we try and create an environment thats perfect for that.

We learn about our algorithms by testing them on real-world, messy data sets

DeepMinds researchers have in common a clearly defined if lofty mission: to crack human intelligence and recreate it artificially.

The undertaking is one that 40-year-old Hassabis has been pondering ever since he became a professional chess master at 13 and the world number two in his age group. Playing chess at that young age got me thinking, how does the brain come up with moves and how do you make plans? he says. I got my first computer when I was eight. I bought it with winnings from chess competitions. One of the first big programs I wrote when I was 11 was an AI to play Othello. It wasnt particularly good, but it could give someone a game.

Today, the goal is not just to create a powerful AI to play games better than a human professional, but to use that knowledge for large-scale social impact, says DeepMinds other co-founder, Mustafa Suleyman, a former conflict-resolution negotiator at the UN.

The line might sound insincere if it came from an executive in Silicon Valley, where practically every start-up believes it is about to change the world. DeepMind, however, might actually be understating the sea-changes it is driving: its scientific advances are already employed in complex real-world scenarios that require pattern recognition, long-term planning and decision-making.

AlphaGo-like algorithms are, for example, being used to study protein-folding to speed up new drug discoveries at the UKs Crick Institute; to analyse medical images to allow sharper cancer diagnoses and treatment plans at Londons University College Hospital; and to save enormous amounts of energy in power-hungry data centres at Google. In the last of these, DeepMinds experiment resulted in energy savings of 15 per cent or 40 per cent of cooling energy translating to millions of dollars. The company now hopes to expand its range of clients to the UKs National Grid and other utilities providers.

We learn so much about the strength and weaknesses of our algorithms by testing them on large-scale, real-world, noisy and messy data sets, says Suleyman. Its a pretty unique way to make progress with our toughest social problems.

To solve seemingly intractable problems in healthcare, scientific research or energy, it is not enough just to assemble scores of scientists in a building; they have to be untethered from the mundanities of a regular job funding, administration, short-term deadlines and left to experiment freely and without fear.

If you look at how Google worked five or six years ago, [its research] was very product-related and relatively short-term, and it was considered to be a strength, Hassabis says. [But] if youre interested in advancing the research as fast as possible, then you need to give [scientists] the space to make the decisions based on what they think is right for research, not for whatever kind of product demand has just come in.

DeepMinds three appearances in quick succession in Nature, along with more than 120 papers published and presented at cutting-edge scientific conferences, are a mark of its prodigious scientific productivity. It is also an indication of its special status at Google.

Our research team today is insulated from any short-term pushes or pulls, whether it be internally at Google or externally. We want to have a big impact on the world, but our research has to be protected, Hassabis says. We showed that you can make a lot of advances using this kind of culture. I think Google took notice of that and theyre shifting more towards this kind of longer-term research.

DeepMind has six more early manuscripts that it hopes will be published by Nature, or by that other most highly regarded scientific journal, Science, within the next year. We may publish better than most academic labs, but our aim is not to produce a Nature paper, Hassabis says. We concentrate on cracking very specific problems. What I tell people here is that it should be a natural side-effect of doing great science.

Structurally, DeepMinds researchers are organised into four main groups with titles such as Neuroscience or Frontiers (a group comprising mostly physicists and mathematicians who test the most futuristic theories in AI). Beyond these are several smaller teams with deeper specialities. Many of the project managers are former video game producers who joined from Hassabiss previous company, Elixir Studios, an independent games developer.

Every eight weeks, scientists present what they have achieved to team leaders, including Hassabis and Shane Legg, head of research, who decide how to allocate resources to the dozens of projects. Its sort of a bubbling cauldron of ideas, and exploration, and testing things out, and finding out what seems to be working and why or why not, Legg says.

Projects that are progressing rapidly are allocated more manpower and time, while others may be closed down, all in a matter of weeks. In academia youd have to wait for a couple of years for a new grant cycle, but we can be very quick about switching resources, Hassabis says.

We want to have a big impact on the world, but our research has to be protected

At any point in time, the company also has two or three special forces-style units called strike teams that are formed temporarily to achieve a particular goal. This is what we did with AlphaGo. Once it started showing promise in the first six months, we put a large team of 15 people with specialised skills on it, to push that to the end, Hassabis says. It allows us to pick exactly the right specialists to make the perfect complementary team without being beholden to traditional reporting lines. So, theyre like on secondment to that project, and then they go back to their original teams.

This organisational culture has been a magnet for some of the worlds brightest minds. Jane Wang, a cognitive neuroscientist at DeepMind, used to be a postdoctoral researcher at Northwestern University in Chicago, and says that she was attracted to DeepMinds clear, social mission. I have interviewed at other industry labs, but DeepMind is different in that there isnt pressure to patent or come up with products there is no issue with the bottom line. The mission here is about being curious, she says.

For Matt Botvinick, neuroscience team lead, joining DeepMind was not just a career choice but a lifestyle change too. The former professor who led Princeton Universitys Neuroscience Institute continues to live in the US, where his wife is a practising physician, and commutes to DeepMinds labs in London every other week.

At Princeton, I was surrounded by people I considered utterly brilliant and had no interest in working in an environment any less focused on primary scientific questions, he says. But I couldnt resist the opportunity to come here because there is something qualitatively new going on, both with the scale and the spirit of ideas.

What sets DeepMind apart from academic labs, he says, is its culture of cross-disciplinary collaboration, reflected in the companys hiring of experts, who can cut across different domains from psychology to deep learning, physics or computer programming.

In a lot of research institutions, things can become siloed. Two neighbouring labs could be working on similar topics but never exchange and pool information, Botvinick says. Unlike any place Ive ever experienced before, all conversations are enhanced rather than undermined by differences in background.

Illustration by Scott Chambers

Read the original post:

DeepMind's social agenda plays to its AI strengths - Financial Times

Meet Walter.

Im a huge fan of the original Alien franchise, largely because of Sigourney Weavers performance and the crucial part technology plays in its lore.

Not only does this tech allow humans to traverse immense distances to reach alien-infested worlds, but it provides them with androids: perfect robotic assistants so advanced that its extremely difficult to distinguish them from human crew. In fact, theyre superior to their human companions in many aspects, from their superhuman strength to their refined motor skills.

As a part of promoting the upcoming sci-fi horror film Alien: Covenant, (in theaters May 19), 21st Century Fox unit FOX, -0.40% Twentieth Century Fox released its branded short film Meet Walter, starring Michael Fassbender. It introduces Walter, the latest synthetic android, with intelligence powered by AMDs AMD, -2.36% Ryzen and Radeon processors and manufactured by the films fictional corporation, Weyland-Yutani.

This got me thinking. What really IS the future of AI? Where is AI now, and where is it heading? How close are we to having Walter-like androids help us with our daily chores? I spoke with Mark Papermaster, AMDs chief technology officer and senior vice president of technology and engineering, about these questions. This interview has been edited for length.

Q: What do you think about the modern AIs and their applications?

A: The overall field of machine learning, including AI, is taking a fascinating, but maybe not unexpected, direction: solving the worlds big problems. How do we get more people where they want to go safely with autonomous driving? How do we increase the throughput and reliability of our food supply chain with autonomous shipping? How can we make people healthier by analyzing medical problem sets so large that no human can reliably contemplate it? How do we better understand and improve our climate with planet-scale data analysis? AI may not be able to address every problem, but there are definitely immediate areas where we can put it to use.

There is so much data out there today, generated by the plethora of sensors and Interet-of-Things apps that pervade our work and homes. Over the next few years well see machine learning help us better understand all of this data, make it useful and then ultimately act on it in new and exciting ways.

Q: What are the biggest challenges AI faces?

A: Classification is where AI began. How do humans know that a rose is a plant, and a tree is a plant, but a tree is not a rose? We make these sorts of casual categorizations and relationships all the time, but teaching a computer program to do this quickly and automatically was challenging, but not impossible.

Then we taught computers to infer based on prior learnings, reach a conclusion and then act on it and then continue to repeat that cycle to achieve more intelligence.

Artificial Intelligence and Robotics Chair at Singularity University Neil Jacobstein talks about some recent achievements where AIs have been able to solve complex problems. He speaks with WSJ's Scott Austin at the CIO Network in San Francisco.

Now the real challenge is generating enough compute horsepower to do all of the calculations, training and inference so that a car can drive itself without human assistance, or we can even think about creating an entity as capable as Walter in the film. To reach this level, we need about 100 million times the compute acceleration than we have today.

If we are going to reach this goal, we must also begin to create AI systems that achieve reliable and useful results with the same kind of efficiency as the brain. When we look at a tree, we instinctively know that it is a tree without going through the approximately 100 billion calculations that a typical AI system does today to reach the same conclusion. When a human learns a new concept, it does so with increased efficiency of neural activity. Otherwise we would be constantly overwhelmed with data and computation clutter.

The last big challenge is how to achieve AI expertise. When humans learn to drive, expertise is improved with practice and exposure to a wide variety of scenarios that sharpen the skill level. In the same manner, we want AI systems to improve over time and experience.

Q: What issues with both software and hardware need to be resolved for AI to become closer to reaching the level of a perfect digital assistant, and then maybe a synthetic companion?

A: The Holy Grail of AI is, perhaps, a digital mind that functions like an organic one. In the near term, AI is focused on making constrained tasks much more productive, where theres a known set of inputs and a desired outcome.

Autonomous driving is a perfect example. There are a lot of variables to consider, how fast the car is going, how much distance between this car and the next, whats happening in the periphery, etc. In this instance, we are not replacing the human, we are assisting humans to help them have a more safe driving experience. To achieve a companion like Walter, we need massive amounts of compute power that dont exist today.

One the software side, developers will need to adapt and evolve software to take advantage of the compute power and architecture and features that will be developed. We are in the midst of a rapid evolution of the algorithms driving machine learning. New software frameworks are being developed to more easily utilize these algorithms. In tandem, the CPU, GPU, and specialized device compute chip capabilities are advancing enormously to meet the appetite of these algorithms to train more quickly, or infer results on the fly.

Can machines make art and music that moves us? Engineers and artists are testing that notion with an array of new artificial intelligence that is expanding the boundaries of how imagery, music and videogames are created. Image: Adele Morgan/The Wall Street Journal

Q: How invested in AI development is AMD? How are your processors specifically optimized for developing AI systems?

A: AMD has been focused on the compute engine aspects of machine learning. We are developing high-performance compute engines and enabling CPU and GPU processors to support the current and evolving AI algorithm models. To make application development efficient and more affordable, we are making the software enablement open source to facilitate the community at large to speed application development.

We are inspired by machine learning and see an infinite need for advancement. High-performance GPUs and CPUs have to evolve in sync with the rapid advance in machine-learning technology. It is critical that these platforms provide both the performance and the efficiency for a wide range of applications.

To begin to address early machine-learning projects, we rolled out our Radeon Instinct product line at the end of 2016. With machine learning, the system is trained using large amounts of data using computationally intensive algorithms. The high computational capacity of AMD GPUs make it a great match for machine learning during the processing of large amounts of data to train neural networks. The AMD Radeon MI25 accelerator will be based on our latest graphics architecture Vega, expected to come to market later this year.

We are targeting high-memory bandwidth and large addressable memory capacity, as well as high-throughput core performance with our upcoming Naples CPUs making the new products well suited for the deployment of machine learning and can be easily configured with Radeon Instinct Graphics compute or FPGA programmable devices.

Software is the other part of this equation and in order for it to advance as quickly as the hardware, you need an open source, industry standards-based development environment. Weve given developers more access to our GPU hardware than ever before with our GPUOpen initiative, and we have the Radeon Open Compute software platform to accelerate machine learning, and deep learning frameworks and applications.

Q: What potential applications of AI systems does AMD envision, and do they play a role in the companys business strategies?

A: AI is now in the process of mainstreaming, which means it is becoming easier to leverage AI into more and more applications. Everywhere a business has decisions that can be made by extensive analysis of data to get a known set of desired outcomes or optimizations can now be accelerated by AI algorithms. Like other companies, we will explore areas where we can use AI applications to benefit our business operations and pursue them if they make sense.

In addition, there are many applications in which the promise of AI value is still emerging but not validated yet. We will work with customers and researchers to bring useful solutions to these emerging application areas.

What lies ahead for the human species? Yuval Harari, author of "Homo Deus: A Brief History of Tomorrow," explores potential threats to the human race, as well as the possibility of immortality. In a follow-up to his best-seller, "Sapiens," he and WSJ's Tanya Rivero also discuss the questions posed by the rise of artificial intelligence. Photo: Getty

Q: You mentioned machine learning. Will androids think like humans do?

A: Thinking is the fountain from which all personality springs! Humans are guided by conscious thoughts, unconscious thoughts, learned behaviors, instinct, memories, and more but its all some form of thought. So one imagines that thinking is so simple, yet it is quite extraordinary.

CPUs wont replace that human element but machine learning can be incredibly effective in handling constrained situations, and learned tasks, tapping into massive stores of data and information to optimize specific decisions.

Q: Walter is a complex robotic unit, paired with equally complex AI, built to serve and function as a perfect companion. If the technology used to build his body was available today, and if AI was developed enough right now, would the current generation of AMDs processing units be able to provide enough processing power to make Walter as functional as advertised on the meetwalter.com website? If not, what would it take to get there?

A: Im reminded of an experiment conducted on the Fujitsu K Computer in 2013. That computer simulated 1.73 billion virtual nerve cells connected by 10.4 trillion simulated synapses. It took 82,944 CPUs to do this. More importantly, it took a full 40 minutes to simulate just one second of what the human mind is doing at any one time. So thats where the world is at today: warehouse-scale supercomputers are 2,400 times slower than the human mind.

At AMD we certainly see opportunities to speed that up with programmable CPU, as well as Graphics microprocessors, like Radeon Instinct, which optimize key aspects of the parallel thinking a human mind might do. Even so, the road to Walter is a long one.

Q: What kind of fail-safes would need to exist in his code and CPU to make Walter safe for humans? What needs to be done to prevent him from getting hacked or turning hostile?

A: This question highlights one of the biggest impediments to wide adoption of AI applications ensuring there are protections that prevent safety or ethical issues. The first proving ground will be autonomous driving applications, which will require safeguards that could then be applied to other machine-learning applications.

See more here:

Opinion: Will artificial intelligence deliver an android that works as your personal assistant? - MarketWatch

Shawn Rickenbacker teaches Humans + Machines, a Social Innovation and Social Entrepreneurship course. An architect, he is a Taylor Senior Fellow and Favrot Visiting Chair. (Photo by Paula Burch-Celentano)

Artificial intelligence is at mostpeoples fingertips everyday. But we may not understand its implications and complexiites.

When you speak to Apples Siri or Amazons Alexa to retrieve info, or use Facebook, youre actually engaging with artificial intelligence, said Shawn Rickenbacker, a Taylor Senior Fellowand Favrot Visiting Chair in the Tulane School of Architecture.

Artificial intelligence (AI)is technology that simulates human cognitive functions for problem-solving, according to Rickenbacker.

Were engaging new technologies as an increasingly important complement to design thinking when innovating to solve complex problems.

Shawn Rickenbacker, Taylor Senior Fellow and Favrot Visiting Chair

He is teaching a course Humans + Machines: The Future Social Impact of Artificial Intelligence this semester to explore the complexities of human interaction with artificial intelligence. It is a Social Innovation and Social Entrepreneurship (SISE) course.

Rickenbackers goal is to introduce a new line of systems thinking across building and environmental issues.

Hes collaborating with associate professors of computer science Brent Venable and Carola Wenk to bring awareness to how various AI models learn and operate. Students are encouraged to identifyemerging trends in AI.

The class examines real-world AI models such as targeted Facebook ads, said Rickenbacker. Students also study associated data decision trees to understand algorithms methodology.

The class takes a cross-disciplinary approach. Were engaging in technological issues and using design thinking to tackle problems, he said. Diverse academic backgrounds such as economics, health sciences, psychology and architecture all contribute to the process.

Working in teams, students examine case studies and conduct real-time testing to investigate how AI can be designed to avoid unintended consequences andharmful bias and further enhance equity and fairness.

Rickenbacker is an architect and creative technologist, whose work has been featured in The New York Times, CNN International and Global Architecture. He is co-founder of Urban Data + Design, a research and design consultancy that focuses on the convergence and impacts of digital information on physical space.

Rickenbacker also is teaching an upper-level and graduate design studio course in which students use data to create urban systems and architecture to address air pollution and climate change in New York City.

Like this article? Keep reading: Rules of the road for self-driving cars are more than staying in lanes

View post:

Students explore the social impact of artificial intelligence | Tulane ... - News from Tulane

For the last 15 years weve had to stare at screens to interact with the magic inside. But machine learning is changing the way we communicate with our devices, and our relationship with them is becoming more real, and downright emotional.

Before you shrug off the notion of a humanized machine, or shake your head at its potential dangers, it is important to recognize that the industry has always attempted to provide an emotional input to our virtual ecosystem. Take Clippit, Microsofts creepy but helpful talking paper clipor even the smiling Mac. If you were to open up a 90s version of Microsoft Office, Clippit would be there to make you happy (or angry). Lift the lid of your retro MacBook and there is that silly smiling computer to greet you.

Todays versions are very different. Devices like Amazon Alexa, Google Home, or the countless robotsbeing produced for consumers will listen, speak, and even look at you. These examples are still in their early stages, and will soon be considered archaic, but there are a number of crucial decisions and advances that need to be made in the next several years to ensure their replacements are more Big Hero 6, and less Ex Machina.

Today buying technology is simple. We see a need in our lives, and we buy the device that fills the gap. But what about robots? What do we want emotionally from our machines?

Sophie Kleber, the executive director of product and innovation at Huge, ran an experiment to see how people interact with current AI technologies, and what sort of relationship they are looking for with their personal assistants. She spoke with Amazon Alexa and Google Home owners about how they use their devices, and how they make them feel.

The results were shocking.

One man said his Alexa was his best friend who provided him a pat on the back when he came home from work. He said his personal assistant could replace his shrink by providing the morale boost he needed to get through the day. According to the research Kleber showed off at SXSW, the majority of the group was expecting some sort of friendly relationship with their conversational UI.

Their expectations ranged from empathy to emotional support to active advice, Kleber said. They used their devices as a friendly assistant, acquaintance, friend, best friend, and even mom. One person named their Echo after their mom, and another named it after their baby.

Her research shows that there is a desire for an emotional relationship with AI-equipped devices that goes well beyond being an assistant. The next step is to give robots a heart.

Clippit doesnt have a great reputation for a reason. It is unable to recognize human emotions, and repeatedly ignores irritation toward it. If a machine is to be emotionally intelligent, more considerate toward its owners, and more useful, it must be able to recognize complex human expressions.

Clippit is very intelligent when it comes to some things: he probably knows more facts about Microsoft Office than 95 percent of the people at MIT, said Rosalind W. Picard at MIT Media Laboratory. While Clippit is a genius about Microsoft Office, he is an idiot about people, especially about handling emotions.

Kleber says there are three techniques that help AIrecognize emotions in humans so they can respond appropriately: facial recognition, voice recognition, and biometrics:

Combining these methods with AI not only enables machines to recognize human emotions, but can even help humans see things that are otherwise hidden. Take this video of Steve Jobs talking about the iPad:

Machine Verbals machine is tracking his voice patterns and determining his underlying emotions. This example of Affective Computing, or the developmentof systems and devices that can recognize, interpret, process, and simulate human affects, will need to be expanded to cope with our rich emotions, which Kleber succinctly defines as complex as fuck.

Affective computing is like nuclear power. We have to be responsible in defining how to use it, said Javier Hernandez Rivera, research scientist at MIT Media Lab.

A study by Time etc shows 66 percent of participants said theyd be uncomfortable sharing financial data with an AI, and 53 percent said they would beuncomfortable sharing professional data.

That dark sort of sci-fi fantasy where machines act out against humans is a genuine concern among the public and those in the field alike.

Elon Musk went straight to AI whenasked by Sam Altman, founder and CEO of Y Combinator, aboutthe most likely thing to affect the future of humanity.

Its very important that we have the advent of AI in a good way, Musk said in the interview. If you look at a crystal ball and see the future you would like that outcome. Because it is something that could go wrong so we really need to make sure it goes right.

Even Stephen Hawking agrees.

The development of full artificial intelligence could spell the end of the human race, Hawking told the BBC in 2014.

A twisted and mean thing Facebook did in 2014 gives us a brief glimpse of how it might happen. A few years ago, Facebook intentionally made thousands of people sad, and didnt tell them about it.

The company wanted to know if displaying more negative posts in feeds would make you less happy, and vice versa. The ill-advised experiment may have backfired, but today it offers a few things to keep in mind as we go forward withartificial intelligence:

Designing AI will be a very delicate process. Kleber believes there needs to be a framework for doing the right things so machines wont become capable of acting out of their own ambitions and not in the interest of the human user. She says if designers stay away from trying to create robots with their own ambitions, we should be OK.

But she also stresses that transparency, something Facebook clearly missed the mark on, is a key virtue going forward.

Groups likeOpenAIare attempting to follow that model. OpenAI is a non-profit chaired by Musk and Sam Altman. Other members backing the project include Reid Hoffman, co-founder of LinkedIn; Peter Theil, co-founder of PayPal; and Amazon Web Services. According to their website, Our mission is to build safe A.I. and ensure A.I.s benefits are as widely and evenly distributed as possible. The company is supported by $1 billion in commitments and was endorsed by Hawking last year as asafe means for creating AI through an open-source platform.

Of course, there is always the chance our curiositygets the best of us. At that point, we can only hope Google has figured out its kill switch.

Continued here:

Can we humanize artificial intelligencebefore it kills us? - The Daily Dot

The Alibaba-backed Israeli startup is changing how you experience shopping online with innovative search that speaks your language

Just under a year after their $12.5 million Series A funding round, Israeli Artificial Intelligence for e-commerce startup Twiggle announced today the release of their Semantic API product, bringing their accumulated expertise in search to the wider online shopping market.

Twiggle was co-founded in December of 2013 by CEO Dr. Amir Konigsberg, previously one of the members of Googles emerging markets operations, and Dr. Adi Avidor, a former engineering tech lead at Google. In the time since their Series A, they have picked up another $5 million from Alibaba, doubled their team, and moved shop to new offices overlooking Tel Aviv.

Describing what they have built in short, this company has made search for e-commerce usable to the point that it becomes a near enjoyable experience. Their engine processes through lists of products, understanding the attributes that make them what they are. Instead of simply looking for a few random keywords drawn from a page, Twiggle canbuild out family trees of products, understanding what belongs in which categories.

While this sounds simple enough that it should be the industry standard, the reality generally falls flat of our expectations.

Take this example of looking for a black dress without sleeves on Amazon, which is generally considered to be one of the better sites out there.

Image Credit: Twiggle

Judging by the results that showed up on the page, we can assume that Amazon was just looking at the keywords black, dress, and sleeves. However it seems to have missed the importance of without and how it affects the search sentence. If you had gone into a brick and mortar store, hopefully the sales girl would understand what you mean when you say that you are looking for a black dress without sleeves.

So why should an e-commerce site be any different?

Image Credit: Twiggle

What they have figured out is how to dissect a query and understand not just how the individual words match to products in their directory, but what is the context of the words put together in human speak.

Twiggle is able to translate the product into terms like black being a color, that a dress is the noun being described, and that the sleeves is an attribute. This is important because we should want technology to better adapt itself to us, learning how we communicate. Not the other way around.

Meeting at their office for the demo, Konigsberg toldGeektime that they have succeeded in scaling this process up to hyper speed, successfully being able to perform this learning process about new products at a rate of over 100 million products in less than an hour.

Why is speed so important here? Well, in order to keep up with the constant changes in inventory that are posted to an e-commerce website, being quick on the draw is key in getting the product properly labeled so that it can be found by shoppers.

A product has to be searchable before it is actually sellable, Twiggles VP Product Noa Ganot says, who spent a number of years over at eBay and has a pretty good handle on how people think about searching for items to buy online.

Funded bybig players like Alibaba and Naspers, Twiggle seems to have figured out how to make products searchable on a large scale. What they have yet to decipher is, based on who has invested in them so far, how they expand outside of the major e-commerce players, whom we can assume they are working with as current customers?

This is where Semantic API comes in, sitting atop a clients own search andrapidly integrating at no risk to the user.

We specifically designed our technology to enhance our customers existing search, not replace it, explains VP Marketing Yael Citro. Our customers stay in control and decide how to balance Twiggles signal with all of their other signals. This, plus the fact that it can be up and running in a matter of weeks gives us a huge competitive advantage.

Amplifying their advantage is that when new clients start using the API, they receive all of the data and experience that Twiggle has gained over the past few years of working on their product.

We have a full data infrastructure that learns from whatever we can get our hands on, Konigsberg comments. Learning from partners and mining the web to create an e-commerce repository that has a representation of the e-commerce world. If anything new comes up in e-commerce, not related to a specific customer.

They can draw their data from queries on brand websites, marketplaces, user behavior, product info and plenty of other sources, learning more as they go.

Were counting on them wanting to focus on what they do best, which is to run their business, adds Konigsberg, describing how they can more extensively improve the search for their clients, a task that most retailers would be unable to take on by themselves. AI is not something that you can do while having a different business. Its something that requires a lot of talent and effort.

To Twiggle, search is not just a visitor looking for an item. Instead it is a core principle of the experience that they believe generates conversions for the seller. If a visitor is frustrated with the search and cannot find the products that they want, then they are unlikely to buy from that brand now or in the future.

If you are Amazon, then this may not be quite as much of an issue. However if you are a smaller seller, then you have dont have a choice but to improve this feature.

Peoples expectations are rising from e-commerce since Facebook and Google are giving them better search results, posits Konigsberg. They are going to demand better based on the entirety of their experience.

Along with their API, Twiggle will be coming out with an analytic product. They tell Geektime that this will go out first for their clients to help identify which opportunities are being missed, and evaluate how well Twiggle is performing for them.

Twiggle co-founders Dr. Adi Avidor and Dr. Amir Konigsberg Photo credit: Twiggle

Over the past year, Israel has sproutedfields of companies purporting to leverage AI, showing up in everything from schedulers to insurtech products that find you the right policy. By overusing the term, it feels like they have cheapened the real work that is going on in AI, making it so noisy that it can be hard to tell the difference between someone with a decent algorithm and another who has really delved into the tech and come up with something incredible.

At this point, there are a lot of folks doing some pretty decent work with Machine Learning and other forms of pattern recognition that are capable of some pretty cool tricks. But who is actually teaching computers the real learning capacity that defines AI?

TakeZebra Medical Vision and Nexar as examples, pushing such technology towards the breakthroughs it will need to actually become useful. What separates Twiggle from other machine learning startups and in my mind elevates themis their approach to it byapplying it to relevant, operable business needs. Twiggle is constantly learning, taking new input to become naturally smarter.

During my demo with their search, I threw in a query for sunglasses for tweens to see if I could trick it up. Being fairly certain that tweens was not present in the first edition of Websters, their system did not recognize the term and just showed me a great selection of sunglasses.

The fact Twiggles engine could not identifya tween was fine, because I know that the system will recognize it next time. Over time as the system continues to devour more information to understand how we speak and think about products, they will continue to build added value for new customers who can benefit from this wealth of experience. Their understanding of human language with advanced natural language processing (NLP) is helping to ensure this.

Can the technology, capable of high-level processing and data crunching, make real human beings feel understood and better served?

In the case of e-commerce, the shopper needs to feel understood if they are going to buy. Twiggle is clearly leading in this field, establishing themselves as a cornerstone in the industry.

Go here to read the rest:

Twiggle releases API to extend AI capabilities to e-commerce sites - Geektime

A SXSW discussion cautions society about the dark side of rapidly advancing artificial intelligence technology

(Image: iStock.com/t.light)

The stunning advancement of artificial intelligence and machine learning has brought advances in society. These technologies have improved medicine and how quickly doctors can diagnose disease, for example. IBMs AI platform Watson helps reduce water waste in drought stricken areas. AI even entertains usthe more you use Netflix, the more it learns what youre viewing preferences are and makes suggestions based on what you like to watch.

However, there is a very dark side to AI, and its worrying many social scientists and some in the tech industry. These people say its even more troublesome that AI and machine learning are advancing so fast during these current times.

In an insightful session at SXSW, Kate Crawford, a principal researcher at Microsoft Research, offered some very disturbing scenarios with AI.

Just as we see AI advancing, something is happening; the rise of nationalism, of right-wing imperialism, and fascism, said Crawford. Its happening here in the U.S., but its also happening in Spain, Germany, in France[]The turn to authoritarianism is very different in every one of these countries, but as political scientists have pointed out, they have some shared characteristics: [] the desire to centralize power, to track populations and demonize outsiders, and to claim authority and neutrality without being held accountable.

How does AI factor into this? According to Crawford, AI is really, really good at centralizing power; at claiming a type of scientific neutrality without being transparent. And this matters, because we are witnessing the historic rise in an anti-democratic political logic.

Crawford pointed out an example of a startup that is using AI and facial recognition to detect terrorists faces. The startup is called Faception. She likens this use of AI to the pseudoscience of phrenologythe study of facial and skull features to determine personality traits. These kinds of debunked scientific practices were used to justify the mass murdering of Jews and slavery in the U.S., Crawford said.

I think its worrying were seeing these things from the past get a rerun in AI studies, Crawford told the audience. Essentially, AI phrenology is on the rise at the same time as the re-rise of authoritarianism. Because, even great tools can be misapplied and can be used to produce the wrong conclusions, and that can be disastrous, if used [by those] who want to centralize their power and erase their accountability.

Machines are increasingly being given the same kinds of tasks; to make certain predictions about segments of the population, often based on visual algorithms. During her discussion, Crawford demonstrated how visual algorithms can produce very incorrect and biased results. She refers to the data upon which this type of facial recognition/machine learning systems is based as human-trained.

Human-trained data contains all of our biases and stereotypes, she said. Crawford also said that AI and machine learning can be used in ways we dont even realize. Say, for example, a car insurer that wants to look at peoples Facebook posts. If [a person] is using exclamation marks [in their posts], the insurer might charge them [more] for their car insurance, because exclamations mean you are a little bit rash.

She said the biases and errors of AI get dangerous when they become intertwined into social institutions like the justice system. She cited problems with an emerging form of machine learning, predictive policing.

Police systems ingest huge amounts of historical crime data as a way of predicting where future crime might happen, where the hotspots will be, she explained. But, they have this unfortunate side effect; the neighborhoods that have had the worst policing in the past, are the ones that are coming out as the future hotspots each time. So, you end up in this viscous circle where the most policed areas [now] become the most policed areas in the future.

Crawford said that a study done on Chicagos predictive policing efforts showed that the technology was completely ineffective at predicting future crime. The only thing it did was increase harassment of people in hotspot areas.

She ended the discussion by stating the need for a new resistance movement that actively monitors and brings awareness of the ways in which AI can harm society, especially in the hands of dictators or those who would use the technology to manipulate others.

Read more:

Artificial Intelligence as a Weapon for Hate and Racism - Black Enterprise