Monthly Archives: May 2017

In BriefAI is becoming more and more ubiquitous, with reports ofadvancements or new applications coming almost daily. How much dowe know about how it thinks, and how are we trying to find outmore? AI as We Understand It

Most of the AI we know today operates on a principle of deep learning: a machine is given a set of data and a desired output, and from that it produces its own algorithm to solve it. The system then repeats, perpetuating itself. This is called a neural network. It is necessary to use this method to create AI, as a computer can code faster than a human; it would take lifetimes to code it manually.

Professor of Electrical Engineering and Computer Science at MIT Tommi Jaakkola says, If you had a very small neural network, you might be able to understand it. But once it becomes very large, and it has thousands of units per layer and maybe hundreds of layers, then it becomes quite un-understandable. We are at the stage of these large systems now. So, in order to make these machines explain themselves an issue that will have to be solved before we can place any trust in them what methods are we using?

1. Reversing the algorithms.Inimage recognition, this involves programming the machine to produce or modify pictures when the computer recognizes a pattern it has learned. Take the example of a Deep Dream modification of The Creation of Adam, where the AI has been told to put dogs in where it recognizes them. From this, we can learn what constitutes a dog for the A.I: firstly, it only produces heads (meaning this is what largely characterizes a dog, according to it) and secondly, the patterns that the computer recognizes as dogs are clustered around Adam (on the left) and God (on the right).

2. Identifying the data it has used. This process of understanding AI gives AI the command to record extracts and highlight the sections of text that it has used according to the pattern it was told to recognize. Developed first by Regina Barzilay, a Delta Electronics Professor at MIT, this type of understanding applies to AIs that search for patterns in data and make predictions accordingly. Carlos Guestrin, a Professor of Machine Learning at the University of Washington, has developed a similar system that presents the data with a short explanation as to why it was chosen.

3. Monitoring individual neurons. Developed by Jason Yosinski, a Machine Learning Researcher at Uber A.I Labs, this involves using a probe and measuring which image stimulates the neuron the most. This allows us to deduce what the AI looks for the most through a process of deduction.

These methods, though, are proving largely ineffective;as Guestrin says, We havent achieved the whole dream, which is where AI has a conversation with you, and it is able to explain. Were a long way from having truly interpretable AI.

It is important to understand how these systems work, as they are already being applied to industries including medicine, cars, finance, and recruitment: areas that have fundamental impacts on our lives. To give this massive power to something we dont understand could be a foolhardy exercise in trust. This is, of course, providing that the AI is honest, and does not suffer from the lapses in truth and perception that humans do.

At the heart of the problem with trying to understand the machines is a tension.If we could predict them perfectly, it would rob AI of the autonomous intelligence that characterizes it. We must remember that we dont know how humans make these decisions either; consciousness remains a mystery, and the world remains an interesting place because of it.

Daniel Dennet warns, though, that one question needs to be answered before AI is introduced: What standards do we demand of them, and of ourselves? How will we design the machines that will soon control our world without us understanding them how do we code our gods?

More here:

We Still Know Very Little About How AI Thinks - Futurism

Robots are great at doing things theyve been shown how to do, but when presented with a novel problem, such as an unfamiliar shape that needs to be gripped, they tend to choke. AI is helping there in the form of systems like Dex-Net, which uses deep learning to let a robotic arm improvise an effective grip for objects its never seen before.

The basic idea behind the system is rather like how we figure out how to pick things up. You see an object, understand its shape and compare it to other objects youve picked up in the past, then use that information to choose the best way to grab it.

Dex-Net doesnt have the advantage of being a living person with eyes and a memory, so its creators gave it more than six million artificial 3D representations of objects and had it work out the best way, theoretically, to pick up each. In real life, the system looks at an object, compares its point cloud to those in its memory and picks what it thinks is the closest fit.

The researchers presented Dex-Net with dozens of objects it hadnt seen before, and its chosen grip only failed one time. That suggests the system is fairly robust despite being trained on synthetic data plus, it comes up with its candidate grip in an average of less than a second.

Dex-Net is the product of Berkeley roboticists, who are set to present the latest version of the system at a conference in July. They also plan to release the data set of objects and point clouds theyve amassed.

Visit link:

This robot arm's AI thinks like we do about how to grab something ... - TechCrunch

Googles AlphaGo the AI developed to tackle the worlds most demanding strategy game is stepping downfrom competitive matches after defeating the worlds best talent. The latest to succumb isGostop-ranked player,Ke Jie, who lost 3-0 in a series hosted in China this week.

The AI, developedby London-based DeepMind, which wasacquired by Google for around $500 million in 2014,also overcomea team of five top playersduring a week of matches. AlphaGofirst drew headlines last year when it beatformer Go world champion Lee Sedol, and theChina event took things to the next level with matches against19-year-old Jie, and doubles with and against other top Go pros.

Challengers defeated,AlphaGohas cast its last competitive stone, DeepMind CEO Demis Hassabis explained.

This weeks series of thrilling games with the worlds best players, in the country where Go originated, has been the highest possible pinnacle for AlphaGo as a competitive program. For that reason, the Future of Go Summit is our final match event with AlphaGo.

The research team behind AlphaGo will now throw their energy into the next set of grand challenges, developing advanced general algorithms that could one day help scientists as they tackle some of our most complex problems, such as finding new cures for diseases, dramatically reducing energy consumption, or inventing revolutionary new materials.

Go is revered as the planets most demanding strategy game, and thats why it made for an ideal field to both develop AI technology and plot machines against humans. Beyond Google, Tencent is among other tech firmsto have unleashed AIs on the game. While it whips up curiosity and attention,the game simple servesas a stepping stone for future plans which is why DeepMind says it is moving on.

Indeed, the British companyhas already made a foray into more practical everyday solutions. Last year, it agreed to a data-sharing partnership with the UKs National Health Service, however the partnership has been criticized for givinga for-profit company access topersonally identifiable health data of around1.6 million NHS patients.The original arrangement remainsunder investigation by the UKs data protection watchdog, the ICO.

Those snafus arent a reflection on the technology itself, however, andHassabis remains bullishon the impact his firm can make.

If AI systems prove they are able to unearth significant new knowledge and strategies in these domains too, the breakthroughs could be truly remarkable. We cant wait to see what comes next, he said.

While AlphaGo is bowing out at the top, it isnt done with Go altogether. DeepMind is planning to publish a final review paper on how the AI developed since its matches with Lee Sedol last year. It is also developinga teaching tool to help newcomers pick up the ropes of the highlycomplicated game, and to enable more experienced handsto learnthe new and innovative moves that Go has introduced. Top players, even Ke Jie himself, studied up on AlphaGos moves andadded someto their arsenal.

Read this article:

After beating the world's elite Go players, Google's AlphaGo AI is retiring - TechCrunch

Is there a more ambivalent word in English than artificial?

Some artificial things are clearly beneficial, such as artificial organs, artificial insemination as a fertility treatment, and artificial sweeteners as an alternative to sugar for weight and diabetes control. But in other contexts, artificial can have a negative connotation. Artificial people. Artificial ingredients. Artificial turf on a baseball field.

Artificial intelligence, meanwhile, manages to straddle both sides of the fence. Its a term that evokes a range of feelings. AI, of course, is already all around us, whether its Apples Siri giving directions, Netflix suggesting movies you might like based on your choices of and reactions to previous films, or Tesla revolutionizing driving with predictive self-driving capabilities.

And AI is penetrating multiple industriesin myriad ways, such as taking some of the guesswork out of manufacturing design, preventing data breaches, smart ad targeting, analyzing structured and unstructured data for medical diagnoses, and sales forecasting.

But while AIs broad and positive impacts excite many people, some view AI as a job killer and a general existential threat to humanity.

Even one of the smartest people on the planet, Stephen Hawking, is ambiguous about the topic,warningthat AI is either the best or the worst thing ever to happen to humanity.

As a sales and marketing executive, Ive been doing a lot of thinking about AI. And not to go all Pollyanna on you, but I believe that in my field, AI wont replace humans it actually will allow us to be more human. To understand why, start with the fact that the manner in which businesses engage with customers is one of the biggest competitive differentiators today. Customers interact with brands across multiple digital touch points from product research to the buying process to ongoing customer care and they expect an experience that feels personalized.

Thanks to the help of AI capabilities that can anticipate buyer behaviors, marketing technology can automate the entire process of reaching out to customers prescribing what type of content to send based on their previous behaviors and actions, as well as identifying the best time to send it.

So AI allows marketers to abandon linear, one-size-fits-all, persona-based customer engagement in favor of a more adaptive, individualized approach one that would be impossible or infeasible for humans to execute at scale.

Thats a big win forartificial intelligence. Another wininvolves sales and marketing reps themselves and the impact AI can have on how they look at and perform their jobs.

While the fear that AI will replace many jobs has some validity if Im a truck driver, for example, I probably have good reason to be wary of self-driving trucks sales and marketing is one of the industries discovering that AI frees employees from mundane tasks and enables them to be more productive and creative.

Even in sales and marketing, the fact remains that AI is a poor substitute for human interaction in certain situations. No matter how much of the buying process takes place in an automated fashion these days, a human-to-human component will still be essential at key junctures.

This means that even as machines take over more work and AI achieves increasingly human-like degrees of sophistication, human emotions such as empathy and creativity are more critical than ever.

And now individuals in sales and marketing get more time and energy to use them.

Despite how data obsessed sales and marketing have become and the growing reliance on AI in aspects of marketing execution, its still up to people to carry out the ultimate goal connecting the brand with other people.

This is why empathy the ability to understand and share the feelings of another is one of the top traits my company looks for in prospective employees.

Remember that sales and marketing have been and always will be exercises in human relations. Artificial intelligence may be playing an increasingly vital supporting role, but the profession remains all about building better, more meaningful person-to-person relationships.

David Satterwhite is the chief revenue officerat Act-On Software, a company that makes a software-as-a-service product for marketing automation.

Above: The Machine Intelligence Landscape This article is part of our Artificial Intelligence series. You can download a high-resolution version of the landscape featuring 288 companies by clicking the image.

See the article here:

How human creativity plays a role in AI - VentureBeat

At Soul Machines, a company that uses artificial intelligenceto create lifelike avatars that respond to human emotion, a fair amountof their work could be considered unsettling to the average person who fears the coming takeover by our AI-robot overlords.

Its a company that pretty much lives in the uncanny valley, that space between fake and real that can creep people out, but thats not usually what happens when people meet BabyX, said Soul Machines founder Mark Sagar.

Instead, he says, when the baby begins to whimper or cry, some respond in human ways, demonstratingwhat appears to be sympathy similar to the kind they maylavish on a human baby.

Ill probably get about 10 or 15 percent of people respond with thats creepy, and others it doesnt bother them at all. Ultimately its about creating an emotional connection and then people jump right into that, he said.

To see which of these two camps you fall into, watch the video below.

Sagar is an associate professor based at the University of Auckland in New Zealand. Hes won an Academy Award for constructing lifelike animated faces for movies like King Kong and Spider-Man 2, work that began at Sagars Laboratory for Animate Technologies to create human movement designed not by actual human movement but byneural networks.There Sagar and staff combine fields like affective computing, bioengineering, theoretical neuroscience, and AI.

Soul Machines makes avatars ranging from a cartoon strawberry for a kids TV show to Nadia, an avatar voiced by actress Cate Blanchett that is able to serve Australian citizens looking for assistance from the National Disability Insurance Scheme.

Other potential applications range from autonomous characters for VR, education, entertainment and gaming, and as a virtual assistants or customer service agents.

Within the coming year to 18 months, Soul Machines plans to open a platform for people to create their own avatars, like a more realistic Bitmoji.

Potential applications of itstech are numerous, but Soul Machines decided to create an AI baby because babies are natural learning machines and as a way to explore the field of social learning, because the companywants to train AI the same way humans raise children.

Its like really looking at the basics of how parents teach children, Sagar said. How does that interaction loop work? Because if we can create that with a computer, weve actually created a very natural way for people to teach computers.

It also helps lower performance expectations for the AI Sagar believes there wont be anything that approaches adult levels of cognition for a long time.

Avatars are made with biologically inspired cognitive models to give the most lifelike interaction as possible. Sagar isnt as concerned about entering uncanny valley as he is focused onavatars establishing a deep connection.

The brain reacts differently to something it perceives to be alive versus something which it perceives to be inanimate, he said. If you ever see a realistic eye looking at you, youre much more likely to respond than if you see a cartoon eye looking at you.

Responses to human emotion are also part of Soul Machines avatars, which can respond to human emotion it sees through cameras that track facial expression.

So should a Soul Machines avatar be in a store window or kiosk, it might look you in the eye. Look away and it could respond with body language meant to convey disappointment or sadness as a way to get your attention. A shop owner could also choose a more humorous or intellectual appeal, or portray a personality associated with their brand.

With biometrics, Soul Machines can remember faces and use AI to determine the best response based on previous interactions.

In time, affective computing couldbe used to create personality profiles that follow you around the world the waycookies follow you on the web across apps, games, virtual reality, and shop windows to develop an understanding of how to best serve your customer service needs (or manipulate you).

Affective computing is technology that can detecthuman emotion, and its being used to serve peopleads in supermarkets but also to improve sales or boardroom performance and, as Cogito does, help understand when veterans with PTSD need help.

In case you needed things to get even more futuristic or sci-fi, Sagar said in the future he may consider combining affective computing, avatars, and AI designed to mimic the tone, style, and word usage of people both alive and dead.

Its an idea that has been part of the popular imagination for some time but is now coming to life in a series of products and projects.

The New Dimensions in Testimony initiative from the University of Southern California interviews Holocaust survivors and makes a hologram of them. When combined with NLP, a person can ask the hologram virtually any question about their life experience; its like a memoir that can talk to you.

Using old chat conversation transcripts,the startup Luka created a neural network after the death of Roman Mazurenko, a close friend of CEO Eugenia Kuyda. A near-identical scenario played out on the TV show Black Mirror when a woman allowed a company to scrape her husbands old emails and put that avatar into a lifelike robot so she could be with him again.

Its a phenomena Sagar calls the creation of digital ghosts and a virtual spirit world.

Id be very interested in combining Luka-type technology with ours and just seeing the implications of that. I think its really a fascinating thing, he said. Once youve built an avatar, plus youve got the transcripts, you essentially created a digital ghost. Essentially were creating this kind of virtual spirit world.

Demos of BabyX 5.0 will begin in late 2017 or 2018, a company spokesperson said.

This post by Khari Johnson originally appeared on VentureBeat.

Tagged with: artificial intelligence

Original post:

Why Soul Machines Made an AI Baby - UploadVR