Robot Brains Catch Humans in 25 Years, Then Speed Right On By

An android Repliee S1, produced by Japan's Osaka University professor Hiroshi Ishiguro, performing during a dress rehearsal of Franz Kafka's "The Metamorphosis." Phototographer: Yoshikazu Tsuno/AFP via Getty Images

We’ve been wrong about these robots before.

Soon after modern computers evolved in the 1940s, futurists started predicting that in just a few decades machines would be as smart as humans. Every year, the prediction seems to get pushed back another year. The consensus now is that it’s going to happen in ... you guessed it, just a few more decades.

There’s more reason to believe the predictions today. After research that’s produced everything from self-driving cars to Jeopardy!-winning supercomputers, scientists have a much better understanding of what they’re up against. And, perhaps, what we’re up against.

Nick Bostrom, director of the Future of Humanity Institute at Oxford University, lays out the best predictions of the artificial intelligence (AI) research community in his new book, “Superintelligence: Paths, Dangers, Strategies.” Here are the combined results of four surveys of AI researchers, including a poll of the most-cited scientists in the field, totalling 170 respondents.

Human-level machine intelligence is defined here as “one that can carry out most human professions at least as well as a typical human.”

By that definition, maybe we shouldn’t be so surprised about these predictions. Robots and algorithms are already squeezing the edges of our global workforce. Jobs with routine tasks are getting digitized: farmers, telemarketers, stock traders, loan officers, lawyers, journalists -- all of these professions have already felt the cold steel nudge of our new automated colleagues. 

Related: How to Keep Your Job When Your Boss Is a Robot

Replication of routine isn't the kind of intelligence Bostrom is interested in. He’s talking about an intelligence with intuition and logic, one that can learn, deal with uncertainty and sense the world around it. The most interesting thing about reaching human-level intelligence isn’t the achievement itself, says Bostrom; it’s what comes next. Once machines can reason and improve themselves, the skynet is the limit.

Computers are improving at an exponential rate. In many areas -- chess, for example -- machine skill is already superhuman. In others -- reason, emotional intelligence -- there’s still a long way to go. Whether human-level general intelligence is reached in 15 years or 150, it’s likely to be a little-observed mile marker on the road toward superintelligence.

Superintelligence: one that “greatly exceeds the cognitive performance of humans in virtually all domains of interest.”

Inventor and Tesla CEO Elon Musk warns that superintelligent machines are possibly the greatest existential threat to humanity. He says the investments he's made in artificial-intelligence companies are primarily to keep an eye on where the field is headed.

“Hope we’re not just the biological boot loader for digital superintelligence,” Musk Tweeted in August. “Unfortunately, that is increasingly probable.”

There are lots of caveats before we prepare to hand the keys to our earthly kingdom over to robot offspring.

• First, humans have a terrible track record of predicting the future. 
• Second, people are notoriously optimistic when forecasting the future of their own industries. 
• Third, it’s not a given that technology will continue to advance along its current trajectory, or even with its current aims.

Still, the brightest minds devoted to this evolving technology are predicting the end of human intellectual supremacy by midcentury. That should be enough to give everyone pause. The direction of technology may be inevitable, but the care with which we approach it is not.

“Success in creating AI would be the biggest event in human history,” wrote theoretical physicist Stephen Hawking, in an Independent column in May. “It might also be the last.”

ORIGINAL: Bloomberg

By Tom Randall 

Nov 10, 2014

OpenTrons: Open-Source Rapid Prototyping for Biology

PLAY
Accelerating research with open, affordable, & easy-to-use lab robots. We empower people to innovate with biotech.

OT.One

This is the OT.One liquid handling robot, the core of OpenTrons' rapid-prototyping platform for biotech. Shown below with a standard single channel micropipette:

This video shows some of the upgrades available to the OT.One -- the OT.MagWash station for magnetic micro-bead washes, a multi-channel micropipette, and on-board camera (see below for details on the Starter Kit and Pro Upgrade).

 PLAY

Accelerate your research! 

Biologists today spend too much time moving around tiny amounts of liquid by hand. Micropipetting is a repetitive, error-prone, time consuming task, and it is slowing down research.

Pipetting got you down?

The OT.One does pipetting for you so you can focus on moving your project forward. Lower your error rate, free up your time, accelerate your research.

Let OpenTrons take care of the pipetting for you!

Testimonial from a cancer researcher.

 PLAY

Synbio Starter Kit! 

Not a scientist? No problem! Our Synbio Starter Kit can get you up and running with synthetic biology!

 PLAY

Together with Synbiota, we put together a kit that lets you run the "hello world" of synthetic biology -- assemble a DNA design from modular genetic blocks, and boot it up in a bacteria to make it glow red under a blacklight!

The Synbiota kit includes four Genomikon DNA Blocks: 

• an Anchor, 
• a Linker,
•  a red florescent protein gene copied from pink choral, and 
• a Cap that completes the plasmid. 

The kit also includes a standard p200 micropipette, and the OT.MagWash DNA assembly station (video above). You can assemble the DNA either by hand or with an OT.One!

OT.One Backer Options

OT.One backers have three options, the Standard Kit, the Getting Started Kit (add +$100 to the backer level), and the Pro Upgrade (add +$500 to the backer level). 

Standard Kit

Robotics:

• Raspberry Pi on board with wifi, BLE. 
• OpenTrons open-source NodeJS software bundle. 
• Industrial grade aluminum chassis. 
• Measures 2ft x 2ft x 2ft, approx. 60lbs. 
• 100 micron XYZ accuracy with the Smoothieboard Open Hardware 5x stepper motor controller. 

Workbench:

• 15 SPE well-plate capacity. 
• 100% aluminum for easy sterilization. 
• Quick-snap mounting system. 

Liquid Handling:

• Single motor mount for standard hand micropipette, single- or multi- channel. 
• Microliter accuracy, every time.

Getting Started Kit - add $100 to an OT.One

Synbiota "Hello World" 

• Run the "hello world" of synthetic biology -- assemble a DNA design, and boot it up into a bacteria to see it glow!
• Four-part Genomikon block DNA kit, Anchor, Linker, red flouecient protein gene copied from a pink choral, and a cap to complete the plasmid. 
• Lab-safe, non-toxic strain of e. coli, to transform and make glow!
• All the reagents and labware you need for the protocol. 

Standard p200 Micropipette

• Pre-fitted with OpenTrons hardware for easy and precise mounting to the OT.One.

OT.MagWash Station

• Automate magnetic micro-bead wash steps for Genomikon DNA assembly. 

Pro Upgrade - add $500 to an OT.One

Dual Pipette Mount:

• Attach a second standard micropipette, single- or multi- channel. 
• Run protocols requiring two different micropipettes in the same job. 

8-Channel Micropipette

• Eight transfers at a time for high-throughput and multiplexed protocols. 
• Pre-fitted with OpenTrons anchor hardware for easy and precise mounting to the OT.One. 

On-Board Camera

• Open for your development! 
• Future applications include job monitoring, auto positioning, colony picking, process confirmation, etc. 

Mix.Bio

• Mix.Bio is the first ever community for the peer-to-peer development of open-source automated biology protocols.

Design, share, and run protocols on Mix.Bio!

This is a video tour of Mix.Bio.

Mix.Bio lets you easily design protocols in the browser. Drag and drop tools, commands, and locations to put together OpenTrons jobs, then share them with your peers. Or, search through protocols others made and download one to run yourself! 

OpenTrons is full Open-Source! 

We stand on the shoulders of giants -- thanks especially to the Smoothieware team for such an awesome motor controller!

Full Open-Source -- Hardware, Software, Science

Friends and Partners

Risks and challenges

Learn about accountability on Kickstarter

Manufacturing is always a challenge, but it is one we are prepared for. We spent the summer in Shenzhen, China, where we solidified a network of suppliers and manufacturers that we know will deliver quality, affordable robots starting in April 2015.

We also avoided parts in the OT.One that require difficult, time-intensive custom tools and molds to make. That makes manufacturing much less complicated, and is one of the main reasons we are so confident we can start shipping to early-bird backers in April.

Also, we are part of Haxlr8r, a hardware accelerator in Shenzhen, and are working with Seeed Studio, the open-hardware facilitator, for manufacturing. Both Hax and Seeed have tons of experience delivering on Kickstarter campaigns, and we are excited to be a part of their awesome network!

Manufacturing is never easy, but we are as prepared as possible to deliver quality robots starting in April!

ORIGINAL: KickStarter

by OpenTrons

Meet Amelia: the computer that's after your job

A new artificially intelligent computer system called 'Amelia' – that can read and understand text, follow processes, solve problems and learn from experience – could replace humans in a wide range of low-level jobs

Amelia aims to answer the question, can machines think? Photo: IPsoft

In February 2011 an artificially intelligent computer system called IBM Watson astonished audiences worldwide by beating the two all-time greatest Jeopardy champions at their own game.

Thanks to its ability to apply 

• advanced natural language processing, 
• information retrieval, 
• knowledge representation, 
• automated reasoning, and 
• machine learning technologies, Watson consistently outperformed its human opponents on the American quiz show Jeopardy.

Watson represented an important milestone in the development of artificial intelligence, but the field has been progressing rapidly – particularly with regard to natural language processing and machine learning.

In 2012, Google used 16,000 computer processors to build a simulated brain that could correctly identify cats in YouTube videos; the Kinect, which provides a 3D body-motion interface for Microsoft's Xbox, uses algorithms that emerged from artificial intelligence research, as does the iPhone's Siri virtual personal assistant.

Today a new artificial intelligence computing system has been unveiled, which promises to transform the global workforce. Named 'Amelia' after American aviator and pioneer Amelia Earhart, the system is able to shoulder the burden of often tedious and laborious tasks, allowing human co-workers to take on more creative roles.

"Watson is perhaps the best data analytics engine that exists on the planet; it is the best search engine that exists on the planet; but IBM did not set out to create a cognitive agent. It wanted to build a program that would win Jeopardy, and it did that," said Chetan Dube, chief executive Officer of IPsoft, the company behind Amelia.

"Amelia, on the other hand, started out not with the intention of winning Jeopardy, but with the pure intention of answering the question posed by Alan Turing in 1950 – can machines think?"

Amelia learns by following the same written instructions as her human colleagues, but is able to absorb information in a matter of seconds. She understands the full meaning of what she reads rather than simply recognising individual words. This involves 

• understanding context, 
• applying logic and 
• inferring implications.

When exposed to the same information as any new employee in a company, Amelia can quickly apply her knowledge to solve queries in a wide range of business processes. Just like any smart worker she learns from her colleagues and, by observing their work, she continually builds her knowledge.

While most ‘smart machines’ require humans to adapt their behaviour in order to interact with them, Amelia is intelligent enough to interact like a human herself. She speaks more than 20 languages, and her core knowledge of a process needs only to be learned once for her to be able to communicate with customers in their language.

Independently, rather than through time-intensive programming, Amelia creates her own 'process map' of the information she is given so that she can work out for herself what actions to take depending on the problem she is solving.

"Intelligence is the ability to acquire and apply knowledge. If a system claims to be intelligent, it must be able to read and understand documents, and answer questions on the basis of that. It must be able to understand processes that it observes. It must be able to solve problems based on the knowledge it has acquired. And when it cannot solve a problem, it must be capable of learning the solution through noticing how a human did it," said Dube.

IPsoft has been working on this technology for 15 years with the aim of developing a platform that does not simply mimic human thought processes but can comprehend the underlying meaning of what is communicated – just like a human.

Just as machines transformed agriculture and manufacturing, IPsoft believes that cognitive technologies will drive the next evolution of the global workforce, so that in the future companies will have digital workforces that comprise a mixture of human and virtual employees.

Amelia has already been trialled within a number of Fortune 1000 companies, in areas such as manning technology help desks, procurement processing, financial trading operations support and providing expert advice for field engineers.

In each of these environments, she has learnt not only from reading existing manuals and situational context but also by observing and working with her human colleagues and discerning for herself a map of the business processes being followed.

In a help desk situation, for example, Amelia can understand what a caller is looking for, ask questions to clarify the issue, find and access the required information and determine which steps to follow in order to solve the problem.

As a knowledge management advisor, she can help engineers working in remote locations who are unable to carry detailed manuals, by diagnosing the cause of failed machinery and guiding them towards the best steps to rectifying the problem.

During these trials, Amelia was able to go from solving very few queries independently to 42 per cent of the most common queries within one month. By the second month she could answer 64 per cent of those queries independently.

"That’s a true learning cognitive agent. Learning is the key to the kingdom, because humans learn from experience. A child may need to be told five times before they learn something, but Amelia needs to be told only once," said Dube.

"Amelia is that Mensa kid, who personifies a major breakthrough in cognitive technologies."

Analysts at Gartner predict that, by 2017, managed services offerings that make use of autonomics and cognitive platforms like Amelia will drive a 60 per cent reduction in the cost of services, enabling organisations to apply human talent to higher level tasks requiring creativity, curiosity and innovation.

IPsoft even has plans to start embedding Amelia into humanoid robots such as Softbank's Pepper, Honda's Asimo or Rethink Robotics' Baxter, allowing her to take advantage of their mechanical functions.

"The robots have got a fair degree of sophistication in all the mechanical functions – the ability to climb up stairs, the ability to run, the ability to play ping pong. What they don’t have is the brain, and we’ll be supplementing that brain part with Amelia," said Dube.

"I am convinced that in the next decade you’ll pass someone in the corridor and not be able to discern if it’s a human or an android."

Given the premise of IPsoft's artificial intelligence system, it seems logical that the ultimate measure of Amelia's success would be passing the Turing Test – which sets out to see whether humans can discern whether they are interacting with a human or a machine.

Earlier this year, a chatbot named Eugene Goostman became the first machine to pass the Turing Test by convincingly imitating a 13-year-old boy. In a five-minute keyboard conversation with a panel of human judges, Eugene managed to convince 33 per cent that it was human.

Interestingly, however, IPsoft believes that the Turing Test needs reframing, to redefine what it means to 'think'. While Eugene was able to immitate natural language, he was only mimicking understanding. He did not learn from the interaction, nor did he demonstrate problem solving skills.

"Natural language understanding is a big step up from parsing. Parsing is syntactic, understanding is semantic, and there’s a big cavern between the two," said Dube.

"The aim of Amelia is not just to get an accolade for managing to fool one in three people on a panel. The assertion is to create something that can answer to the fundamental need of human beings – particularly after a certain age – of companionship. That is our intent."

ORIGINAL: Telegraph

By Sophie Curtis
29 Sep 2014

5 Robots Booking It to a Classroom Near You

IMAGE: ANDY BAKER/GETTY IMAGES

Robots are the new kids in school.

The technological creations are taking on serious roles in the classroom. With the accelerating rate of robotic technology, school administrators all over the world are plotting how to implement them in education, from elementary through high school.

In South Korea, robots are replacing English teachers entirely, entrusted with leading and teaching entire classrooms. In Alaska, some robots are replacing the need for teachers to physically be present at all.

SEE ALSO: 11 Ways to Make Science Fun for Your Kids

Robotics 101 is now in session. Here are five ways robots are being introduced into schools.

1. Nao Robot as math teacher

IMAGE: WIKIPEDIA

In Harlem school PS 76, a Nao robot created in France, nicknamed Projo helps students improve their math skills. It's small, about the size of a stuffed animal, and sits by a computer to assist students working on math and science problems online.

Sandra Okita, a teacher at the school, told The Wall Street Journal the robot gauges how students interact with non-human teachers. The students have taken to the humanoid robotic peer, who can speak and react, saying it's helpful and gives the right amount of hints to help them get their work done.

2. Aiding children with autism

The Nao Robot also helps improve social interaction and communication for children with autism. The robots were introduced in a classroom in Birmingham, England in 2012, to play with children in elementary school. Though the children were intimidated at first, they've taken to the robotic friend, according to The Telegraph.

3. VGo robot for ill children

Sick students will never have to miss class again if the VGo robot catches on. Created by VGo Communications, the rolling robot has a webcam and can be controlled and operated remotely via computer. About 30 students with special needs nationwide have been using the $6,000 robot to attend classes.

For example, a 12-year-old Texas student with leukemia kept up with classmates by using a VGo robot. With a price tag of about $6,000, the robots aren't easily accessible, but they're a promising sign of what's to come.

4. Robots over teachers

In the South Korean town of Masan, robots are starting to replace teachers entirely. The government started using the robots to teach students English in 2010. The robots operate under supervision, but the plan is to have them lead a room exclusively in a few years, as robot technology develops.

5. Virtual teachers

IMAGE: FLICKR, SEAN MACENTEE

South Korea isn't the only place getting virtual teachers. A school in Kodiak, Alaska has started using telepresence robots to beam teachers into the classroom. The tall, rolling robots have iPads attached to the top, which teachers will use to video chat with students.

The Kodiak Island Borough School District's superintendent, Stewart McDonald, told The Washington Times he was inspired to do this because of the show The Big Bang Theory, which stars a similar robot. Each robot costs about $2,000; the school bought 12 total in early 2014.

ORIGINAL: Mashable
BY YOHANA DESTA

Everybody Relax: An MIT Economist Explains Why Robots Won't Steal Our Jobs

Living together in harmony. Photo by Oli Scarff/Getty Images

If you’ve ever found yourself fretting about the possibility that software and robotics are on the verge of thieving away all our jobs, renowned MIT labor economist David Autor is out with a new paper that might ease your nerves. Presented Friday at the Federal Reserve Bank of Kansas City’s big annual conference in Jackson Hole, Wyoming, the paper argues that humanity still has two big points in its favor: People have "common sense,” and they’re "flexible."

Neil Irwin already has a lovely writeup of the paper at the New York Times, but let’s run down the basics. There’s no question machines are getting smarter, and quickly acquiring the ability to perform work that once seemed uniquely human. Think self-driving cars that might one day threaten cabbies, or computer programs that can handle the basics of legal research.

But artificial intelligence is still just that: artificial. We haven’t untangled all the mysteries of human judgment, and programmers definitely can’t translate the way we think entirely into code. Instead, scientists at the forefront of AI have found workarounds like machine-learning algorithms. As Autor points out, a computer might not have any abstract concept of a chair, but show it enough Ikea catalogs, and it can eventually suss out the physical properties statistically associated with a seat. Fortunately for you and me, this approach still has its limits.

For example, both a toilet and a traffic cone look somewhat like a chair, but a bit of reasoning about their shapes vis-à-vis the human anatomy suggests that a traffic cone is unlikely to make a comfortable seat. Drawing this inference, however, requires reasoning about what an object is “for” not simply what it looks like. Contemporary object recognition programs do not, for the most part, take this reasoning-based approach to identifying objects, likely because the task of developing and generalizing the approach to a large set of objects would be extremely challenging.

That’s what Autor means when he says machines lack for common sense. They don’t think. They just do math.

And that leaves lots of room for human workers in the future.

Technology has already whittled away at middle class jobs, from factory workers replaced by robotic arms to secretaries made redundant by Outlook, over the past few decades. But Autor argues that plenty of today's middle-skill occupations, such as construction trades and medical technicians, will stick around, because “many of the tasks currently bundled into these jobs cannot readily be unbundled … without a substantial drop in quality.”

These aren’t jobs that require performing a single task over and over again, but instead demand that employees handle some technical work while dealing with other human beings and improvising their way through unexpected problems. Machine learning algorithms can’t handle all of that. Human beings, Swiss-army knives that we are, can. We’re flexible.

Just like the dystopian arguments that machines are about to replace a vast swath of the workforce, Autor’s paper is very much speculative. It’s worth highlighting, though, because it cuts through the silly sense of inevitability that sometimes clouds this subject. Predictions about the future of technology and the economy are made to be dashed. And while Noah Smith makes a good point that we might want to be prepared for mass, technology-driven unemployment even if there’s just a slim chance of it happening, there’s also no reason to take it for granted.

Jordan Weissmann is Slate's senior business and economics correspondent.

ORIGINAL: Slate

By Jordan Weissmann

"Brain" In A Dish Acts As Autopilot Living Computer

A glass dish contains a "brain" -- a living network of 25,000 rat brain cells connected to an array of 60 electrodes.University of Florida/Ray Carson

downloadable pdf

A University of Florida scientist has grown a living “brain” that can fly a simulated plane, giving scientists a novel way to observe how brain cells function as a network.The “brain” — a collection of 25,000 living neurons, or nerve cells, taken from a rat’s brain and cultured inside a glass dish — gives scientists a unique real-time window into the brain at the cellular level. By watching the brain cells interact, scientists hope to understand what causes neural disorders such as epilepsy and to determine noninvasive ways to intervene.

Thomas DeMarse holds a glass dish containing a living network of 25,000 rat brain cells connected to an array of 60 electrodes that can interact with a computer to fly a simulated F-22 fighter plane.

As living omputers, they may someday be used to fly small unmanned airplanes or handle tasks that are dangerous for humans, such as search-and-rescue missions or bomb damage assessments." We’re interested in studying how brains compute,” said Thomas DeMarse, the UF assistant professor of biomedical engineering who designed the study. “If you think about your brain, and learning and the memory process, I can ask you questions about when you were 5 years old and you can retrieve information. That’s a tremendous capacity for memory. In fact, you perform fairly simple tasks that you would think a computer would easily be able to accomplish, but in fact it can’t.”

While computers are very fast at processing some kinds of information, they can’t approach the flexibility of the human brain, DeMarse said. In particular, brains can easily make certain kinds of computations — such as recognizing an unfamiliar piece of furniture as a table or a lamp — that are very difficult to program into today’s computers.

“If we can extract the rules of how these neural networks are doing computations like pattern recognition, we can apply that to create novel computing systems,” he said.

DeMarse’s experimental “brain” interacts with an F-22 fighter jet flight simulator through a specially designed plate called a multi-electrode array and a common desktop computer. “It’s essentially a dish with 60 electrodes arranged in a grid at the bottom,” DeMarse said. “Over that we put the living cortical neurons from rats, which rapidly begin to reconnect themselves, forming a living neural network — a brain.” The brain and the simulator establish a two-way connection, similar to how neurons receive and interpret signals from each other to control our bodies. By observing how the nerve cells interact with the simulator, scientists can decode how a neural network establishes connections and begins to compute, DeMarse said. When DeMarse first puts the neurons in the dish, they look like little more than grains of sand sprinkled in water. However, individual neurons soon begin to extend microscopic lines toward each other, making connections that represent neural processes. “You see one extend a process, pull it back, extend it out — and it may do that a couple of times, just sampling who’s next to it, until over time the connectivity starts to establish itself,” he said. “(The brain is) getting its network to the point where it’s a live computation device.” To control the simulated aircraft, the neurons first receive information from the computer about flight conditions: whether the plane is flying straight and level or is tilted to the left or to the right.

The neurons then analyze the data and respond by sending signals to the plane’s controls. Those signals alter the flight path and new information is sent to the neurons, creating a feedback system. “Initially when we hook up this brain to a flight simulator, it doesn’t know how to control the aircraft,” DeMarse said. “So you hook it up and the aircraft simply drifts randomly. And as the datacome in, it slowly modifies the (neural) network so over time, the network gradually learns to fly the aircraft.” Although the brain currently is able to control the pitch and roll of the simulated aircraft in weather conditions ranging from blue skies to stormy, hurricane-force winds, the underlying goal is a more fundamental understanding of how neurons interact as a network, DeMarse said. “There’s a lot of data out there that will tell you that the computation that’s going on here isn’t based on just one neuron. 

The computational property is actually an emergent property of hundreds or thousands of neurons cooperating to produce the amazing processing power of the brain.” With José Principe, a UF distinguished professor of electrical engineering and director of UF’s Computational NeuroEngineering Laboratory, DeMarse has a $500,000 National Science Foundation grant to create a mathematical model that reproduces how the neurons compute. Thomas DeMarse, tdemarse@bme.ufl.edu"

ORIGINAL: U of Florida

by Jennifer Viegas  

Nov 27, 2012

Elon Musk: Artificial Intelligence Is 'Potentially More Dangerous Than Nukes'

Google Images

Back in June, Tesla CEO Elon Musk told CNBC that he'd invested in a company called Vicarious that is developing products and services based on artificial intelligence. But that wasn't why Musk got interested. His impetus for backing the firm was instead "to keep an eye on" unforeseen terrifying scenarios where the products began to threaten humanity.

He doesn't appear to have been exaggerating.

In a Tweet last night, Musk said this:

Worth reading Superintelligence by Bostrom. We need to be super careful with AI. Potentially more dangerous than nukes.
— Elon Musk (@elonmusk) August 3, 2014

Hope we're not just the biological boot loader for digital superintelligence. Unfortunately, that is increasingly probable
— Elon Musk (@elonmusk) August 3, 2014

Bostrom is Nick Bostrom, the founder of Oxford’s Future of Humanity Institute. That group recently partnered with a new group at Cambridge, the Centre for the Study of Existential Risk, to study how things like nanotechnology, robotics, artificial intelligence and other innovations could someday wipe us all out, according to PCPro:

At [a] conference, Bostrom was asked if we should be scared by new technology. "Yes," he said, "but scared about the right things. There are huge existential threats, these are threats to the very survival of life on Earth, from machine intelligence – not the way it is today, but if we achieve this sort of super-intelligence in the future," Bostrom said.

"Superintelligence" is set to be published in English next month. In a blurb, Bostrom's colleague Martin Rees of Cambridge says of the work, "Those disposed to dismiss an 'AI takeover' as science fiction may think again after reading this original and well-argued book."

In our recent profile of Vicarious, the firm backed by Musk, we talked to Bruno Olshausen, a Berkeley professor and one of the firm's advisors. He said we are still way too far behind in our understanding of how the brain works to be able to create something that could turn heel.

"Absent a major paradigm shift - something unforeseeable at present - I would not say we are at the point where we should truly be worried about AI going out of control," he told us.

So at a minimum, it sounds like the robot takeover is not imminent.

But it seems like it's something all of us should "keep an eye on."

ORIGINAL: Business Insider

Rob Wile
Aug. 3, 2014

Welcome to the Claytronics Project

Collaborative Research in Programmable Matter Directed by Carnegie Mellon and Intel

This project combines modular robotics, systems nanotechnology and computer science to create the dynamic, 3-Dimensional display of electronic information known as claytronics.

Our goal is to give tangible, interactive forms to information so that a user's senses will experience digital environments as though they are indistinguishable from reality.

Claytronics is taking place across a rapidly advancing frontier. This technology will help to drive breathtaking advances in the design and engineering of computing and hardware systems. 

Our research team focuses on two main projects:

• Creating the basic modular building block of claytronics known as the claytronic atom or catom, and
• Designing and writing robust and reliable software programs that will manage the shaping of ensembles of millions of catoms into dynamic, 3-Dimensional forms.

Realizing the vision of claytronics through the self-assembly of millions of catoms into synthetic reality will have a profound effect on the experience of users of electronic information. This promise of claytronic technology has become possible because of the ever increasing speeds of computer processing predicted in Moore's Law. 

This website will introduce you to the ideas that are driving claytronics, the research team that is working to make it happen, and the hardware and software projects that enable the building of claytronic ensembles.

Development of this powerful form of information display represents a partnership between the School of Computer Sciences of Carnegie Mellon University, Intel Corporation at its Pittsburgh Laboratory and FEMTO-ST Institute. As an integral part of our philosophy, the Claytronics Project seeks the contributions of scholars and researchers worldwide who are dedicating their efforts to the diverse scientific and engineering studies related to this rich field of nanotechnology and computer science. 

To understand the future of claytronics, watch the concept video [.mov] created by Carnegie Mellon's Entertainment Technology Center.

Use the links to the left to see a list of publications, some videos and photos documenting our progress, a partial list of talks we have given, and people working on the project.

ORIGINAL: CMU

Stanford bioengineers create circuit board modeled on the human brain

Stanford bioengineers have developed faster, more energy-efficient microchips based on the human brain – 9,000 times faster and using significantly less power than a typical PC. This offers greater possibilities for advances in robotics and a new way of understanding the brain. For instance, a chip as fast and efficient as the human brain could drive prosthetic limbs with the speed and complexity of our own actions.

The Neurogrid circuit board can simulate orders of magnitude more neurons and synapses than other brain mimics on the power it takes to run a tablet computer.

Stanford bioengineers have developed a new circuit board modeled on the human brain, possibly opening up new frontiers in robotics and computing.

For all their sophistication, computers pale in comparison to the brain. The modest cortex of the mouse, for instance, operates 9,000 times faster than a personal computer simulation of its functions.

Not only is the PC slower, it takes 40,000 times more power to run, writes Kwabena Boahen, associate professor of bioengineering at Stanford, in an article for the Proceedings of the IEEE.

"From a pure energy perspective, the brain is hard to match," says Boahen, whose article surveys how "neuromorphic" researchers in the United States and Europe are using silicon and software to build electronic systems that mimic neurons and synapses.

Boahen and his team have developed Neurogrid, a circuit board consisting of 16 custom-designed "Neurocore" chips. Together these 16 chips can simulate 1 million neurons and billions of synaptic connections. The team designed these chips with power efficiency in mind. Their strategy was to enable certain synapses to share hardware circuits. The result was Neurogrid – a device about the size of an iPad that can simulate orders of magnitude more neurons and synapses than other brain mimics on the power it takes to run a tablet computer.

The National Institutes of Health funded development of this million-neuron prototype with a five-year Pioneer Award. Now Boahen stands ready for the next steps – lowering costs and creating compiler software that would enable engineers and computer scientists with no knowledge of neuroscience to solve problems – such as controlling a humanoid robot – using Neurogrid.

Its speed and low power characteristics make Neurogrid ideal for more than just modeling the human brain. Boahen is working with other Stanford scientists to develop prosthetic limbs for paralyzed people that would be controlled by a Neurocore-like chip.

"Right now, you have to know how the brain works to program one of these," said Boahen, gesturing at the $40,000 prototype board on the desk of his Stanford office. "We want to create a neurocompiler so that you would not need to know anything about synapses and neurons to able to use one of these."

Brain ferment

In his article, Boahen notes the larger context of neuromorphic research, including the European Union's Human Brain Project, which aims to simulate a human brain on a supercomputer. By contrast, the U.S. BRAIN Project – short for Brain Research through Advancing Innovative Neurotechnologies – has taken a tool-building approach by challenging scientists, including many at Stanford, to develop new kinds of tools that can read out the activity of thousands or even millions of neurons in the brain as well as write in complex patterns of activity.

Zooming from the big picture, Boahen's article focuses on two projects comparable to Neurogrid that attempt to model brain functions in silicon and/or software.

One of these efforts is IBM's SyNAPSE Project – short for Systems of Neuromorphic Adaptive Plastic Scalable Electronics. As the name implies, SyNAPSE involves a bid to redesign chips, code-named Golden Gate, to emulate the ability of neurons to make a great many synaptic connections – a feature that helps the brain solve problems on the fly. At present a Golden Gate chip consists of 256 digital neurons each equipped with 1,024 digital synaptic circuits, with IBM on track to greatly increase the numbers of neurons in the system.

Heidelberg University's BrainScales project has the ambitious goal of developing analog chips to mimic the behaviors of neurons and synapses. Their HICANN chip – short for High Input Count Analog Neural Network – would be the core of a system designed to accelerate brain simulations, to enable researchers to model drug interactions that might take months to play out in a compressed time frame. At present, the HICANN system can emulate 512 neurons each equipped with 224 synaptic circuits, with a roadmap to greatly expand that hardware base.

Each of these research teams has made different technical choices, such as whether to dedicate each hardware circuit to modeling a single neural element (e.g., a single synapse) or several (e.g., by activating the hardware circuit twice to model the effect of two active synapses). These choices have resulted in different trade-offs in terms of capability and performance.

In his analysis, Boahen creates a single metric to account for total system cost – including the size of the chip, how many neurons it simulates and the power it consumes.

Neurogrid was by far the most cost-effective way to simulate neurons, in keeping with Boahen's goal of creating a system affordable enough to be widely used in research.

Speed and efficiency

But much work lies ahead. Each of the current million-neuron Neurogrid circuit boards cost about $40,000. Boahen believes dramatic cost reductions are possible. Neurogrid is based on 16 Neurocores, each of which supports 65,536 neurons. Those chips were made using 15-year-old fabrication technologies.

By switching to modern manufacturing processes and fabricating the chips in large volumes, he could cut a Neurocore's cost 100-fold – suggesting a million-neuron board for $400 a copy. With that cheaper hardware and compiler software to make it easy to configure, these neuromorphic systems could find numerous applications.

For instance, a chip as fast and efficient as the human brain could drive prosthetic limbs with the speed and complexity of our own actions – but without being tethered to a power source. Krishna Shenoy, an electrical engineering professor at Stanford and Boahen's neighbor at the interdisciplinary Bio-X center, is developing ways of reading brain signals to understand movement. Boahen envisions a Neurocore-like chip that could be implanted in a paralyzed person's brain, interpreting those intended movements and translating them to commands for prosthetic limbs without overheating the brain.

A small prosthetic arm in Boahen's lab is currently controlled by Neurogrid to execute movement commands in real time. For now it doesn't look like much, but its simple levers and joints hold hope for robotic limbs of the future.

Of course, all of these neuromorphic efforts are beggared by the complexity and efficiency of the human brain.

In his article, Boahen notes that Neurogrid is about 100,000 times more energy efficient than a personal computer simulation of 1 million neurons. Yet it is an energy hog compared to our biological CPU.

"The human brain, with 80,000 times more neurons than Neurogrid, consumes only three times as much power," Boahen writes. "Achieving this level of energy efficiency while offering greater configurability and scale is the ultimate challenge neuromorphic engineers face."

Tom Abate writes about the students, faculty and research of the School of Engineering. Amy Adams of Stanford University Communications contributed to this report.

For more Stanford experts in bioengineering and other topics, visit Stanford Experts.

ORIGINAL: Stanford
BY TOM ABATE
April 28, 2014