Author Archives: Android
Understanding Agent CooperationJoel Leibo, Vinicius Zambaldi, Marc Lanctot, Janusz Marecki, Thore Graepel | Google DeepMind Blog"Recent progress in artificial intelligence and specifically deep reinforcement learning provides us with the tools to look at the problem of social dilemmas through a new lens… we showed that we can apply the modern AI technique of deep multi-agent reinforcement learning to age-old questions in social science such as the mystery of the emergence of cooperation."
Agility Robotics Introduces Cassie, a Dynamic and Talented Robot Delivery OstrichEvan Ackerman | IEEE Spectrum"Agility Robotics, a spin-off of Oregon State University, is officially announcing a shiny new bipedal robot named Cassie. Cassie is a dynamic walker, meaning that it walks much more like humans do than most of the carefully plodding bipedal robots we’re used to seeing… Cassie has some work to do before it’s ready to be hauling groceries up stairs for you, but we’re very much looking forward to watching this robot taking more steps toward robust and dynamic legged locomotion."
How Escape Rooms and Live Theater Are Paving the Way for VRBryan Bishop | The Verge"Cinema has had more than a century to develop its own language of shots, cuts, and transitions, while storytelling in VR is still in its infancy… creators seem to be zeroing in on interactive, experiential moments as one of the key building blocks of VR storytelling. One of Chris Milk’s next projects is a piece set in the Planet of the Apes universe that will lean heavily on AI to drive interactive character performances."
QUANTUM PHYSICS & COSMOLOGY
The Universe Is as Spooky as Einstein ThoughtNatalie Wolchover | The Atlantic "With an experiment described this week in Physical Review Letters—a feat that involved harnessing starlight to control measurements of particles shot between buildings in Vienna—some of the world’s leading cosmologists and quantum physicists are closing the door on an intriguing alternative to 'quantum entanglement.'"
The Army Is Testing Brainwave-Reading Technology for DriversJordan Pearson | MOTHERBOARD "In the study, the researchers developed a series of algorithms to interpret brainwaves based on their previous work, and tried them out on 16 different people strapped into a car in a testing chamber. The test subjects were asked to drive down a long, uniform highway and periodically switch lanes in a simulation. Their brainwaves were recorded during the test, and a video camera kept tabs on them."
FUTURE OF LEARNING
The Techies Who Are Hacking Education by Homeschooling Their KidsJason Tanz | WIRED"'The world is changing. It’s looking for people who are creative and entrepreneurial, and that’s not going to happen in a system that tells kids what to do all day,' Samantha says. 'So how do you do that? Well if the system won’t allow it, as the saying goes: If you want something done right, do it yourself.'"
SILICON VALLEY POLITICS
Companies That Stay Silent on Political Issues Can Pay a Hefty PriceDaniel Korschun | Fast Company"Chevron, Disney, Verizon, GM, Wells Fargo and others have all taken a wait-and-see approach… Such responses are no doubt based on the prevailing wisdom that companies need to stay out of politics… in fact, the opposite may be true: It may be more dangerous to remain silent than to take a political stand… As long as a company is not being deceptive by obfuscating its beliefs, consumers can be surprisingly tolerant of a company that holds an opposing view."
Image source: Shutterstock Continue reading
One day, robots like these will be scampering up your steps to drop off packages Continue reading
Quantum computers promise to crack some of the world’s most intractable problems by super-charging processing power. But the technical challenges involved in building these machines mean they’ve still achieved just a fraction of what they are theoretically capable of.
Now physicists from the UK have created a blueprint for a soccer-field-sized machine they say could reach the blistering speeds that would allow them to solve problems beyond the reach of today’s most powerful supercomputers.
The system is based on a modular design interlinking multiple independent quantum computing units, which could be scaled up to almost any size. Modular approaches have been suggested before, but innovations such as a far simpler control system and inter-module connection speeds 100,000 times faster than the state-of-the-art make this the first practical proposal for a large-scale quantum computer.
“For many years, people said that it was completely impossible to construct an actual quantum computer. With our work we have not only shown that it can be done, but now we are delivering a nuts and bolts construction plan to build an actual large-scale machine,” professor Winfried Hensinger, head of the Ion Quantum Technology Group at the University of Sussex who led the research, said in a press release.
The technology at the heart of the individual modules is already well-established and relies on trapping ions (charged atoms) in magnetic fields to act as “qubits,” the basic units of information in quantum computers.
While bits in conventional computers can have a value of either 1 or 0, qubits take advantage of the quantum mechanical phenomena of superposition, which allows them to be both at the same time.
As Elizabeth Gibney explains in Nature, this is what makes quantum computers so incredibly fast. “The set of qubits comprising the memory of a quantum computer could exist in every possible combination of 1s and 0s at once. Where a classical computer has to try each combination in turn, a quantum computer could process all those combinations simultaneously.”
In a paper published in the journal Science Advances last week, researchers outline designs for modules containing roughly 2,500 qubits and suggest interlinking thousands of them together to create a machine containing two billion qubits. For comparison, Canadian firm D-Wave, the only commercial producer of quantum computers, just brought out its latest model featuring 2,000 qubits.
This is not the first time a modular system like this has been suggested, but previous approaches have recommended using light waves traveling through fiber optics to link the units. This results in interaction rates between modules far slower than the quantum operations happening within them, putting a handbrake on the system’s overall speed. In the new design, the ions themselves are shuttled from one module to another using electrical fields, which results in 100,000 times faster connection speeds.
The system also has a much simpler way of controlling qubits. Previous designs required lasers to be carefully targeted at each ion, an enormous engineering challenge when dealing with billions of qubits. Instead, the new system uses microwave fields and the careful application of voltages, which is much easier to scale up.
The researchers concede there are still considerable technical challenges to building a device on the scale they have suggested, not to mention the cost. But they have already announced plans to build a prototype based on the design at the university at a cost of £1-2 million.
“While this proposal is incredibly challenging, I wish more in the quantum community would think big like this,” Christopher Monroe, a physicist at the University of Maryland who has worked on trapped-ion quantum computing, told Nature.
In their paper, the researchers predict their two billion qubit system could find the prime factors of a 617-digit-long number in 110 days. This is significant because many state-of-the-art encryption systems rely on the fact that factoring large numbers can take conventional computers thousands of years. This is why many in the cybersecurity world are nervous about the advent of quantum computing.
These researchers aren’t the only ones working on bringing quantum computing into the real world, though. Google, Microsoft and IBM are all developing their own systems, and D-Wave recently open-sourced a software tool that helps those without a background in quantum physics program its machines.
All that interest is due to the enormous potential of quantum computing to solve problems as diverse and complex as developing drugs for previously incurable diseases, devising new breeds of materials for high-performance superconductors, magnets and batteries and even turbo-charging machine learning and artificial intelligence.
"The availability of a universal quantum computer may have a fundamental impact on society as a whole,” said Hensinger. “Without doubt it is still challenging to build a large-scale machine, but now is the time to translate academic excellence into actual application, building on the UK's strengths in this ground-breaking technology.”
Image Credit: University of Sussex/YouTube Continue reading
I’m putting out a call for brilliant entrepreneurs who want to enroll in Singularity University’s Global Solutions Program (GSP).
The GSP is where you’ll learn about exponentially growing technology, dig into humanity’s Global Grand Challenges (GGCs) and then start a new company, product or service with the goal of positively impacting 1 billion people within 10 years.
We call this a “ten-to-the-ninth (10⁹+)” company.
This post is about who should apply, how to apply and the over $1.5 million in scholarships being provided by Google for entrepreneurs.
SU’s GSP program runs from June 17, 2017 until August 19, 2017.
Applications are due: February 21, 2017.
What is Singularity University?
Eight years ago, Ray Kurzweil and I cofounded Singularity University to search the world for the most brilliant, world-class problem-solvers, to bring them together, and to give them the resources to create companies that impact billions.
The GSP is an intensive 24/7 experience at the SU campus at the NASA Research Center in Mountain View, CA, in the heart of Silicon Valley.
During the nine-week program, 90 entrepreneurs, engineers, scientists, lawyers, doctors and innovators from around the world learn from our expert faculty about infinite computing, AI, robotics, 3D printing, networks/sensors, synthetic biology, entrepreneurship, and more, and focus on building and developing companies to solve the global grand challenges (GGCs).
GSP participants form teams to develop unique solutions to GGCs, with the intent to form a company that, as I mentioned above, will positively impact the lives of a billion people in 10 years or less.
Over the course of the summer, participants listen to and interact with top Silicon Valley executive guest speakers, tour facilities like GoogleX, and spend hours getting their hands dirty in our highly advanced maker workshop.
At the end of the summer, the best of these startups will be asked to join SU Labs, where they will receive additional funding and support to take the company to the next level.
Requirements for Admission
You’re an accomplished entrepreneur, technologist, scientist, or change agent with expertise in your chosen field(s).
You have a background in or a passion for addressing climate change or are working to address any of the direct or indirect effects of climate change, such as governance, security, sustainable water supplies, and alternative food or energy sources.
You’re already considering how to leverage exponential technologies to create positive impact in the world.
You’re committed to the long-term journey of launching an impact-focused startup.
What’s the Cost?
I am pleased to announce that thanks to a wonderful partnership with Google, all successful applicants will be fully subsidized by Google to participate in the program.
In other words, if accepted into the program, the GSP is free.
Are You Ready to Change the World?
The Global Solutions Program (GSP) is SU’s flagship program for innovators from a wide diversity of backgrounds, geographies, perspectives, and expertise. At GSP, you’ll get the mindset, tools, and network to help you create moonshot innovations that will positively transform the future of humanity. If you're looking to create solutions to help billions of people, we can help you do just that.
Key program dates:
February 21, 2017: Submission deadline for direct applications
April 14, 2017: Acceptance notification
June 17, 2017: GSP participants arrive at SU
This program will be unlike any we've ever done—and unlike any you've ever seen.
If you feel like you meet the criteria, apply now (click here).
Applications close February 21nd, 2017.
If you know of a friend or colleague who would be a good fit for this program, please share this post with them and ask that they fill out an application. Continue reading
Accelerating technology has been creating a lot of worry over job loss to automation, especially as machines become capable of doing things they never could in the past. A recent report released by the McKinsey Global Institute estimated that 49 percent of job activities could currently be fully automated—that equates to 1.1 billion workers globally.
What gets less buzz is the other side of the coin: automation helping to create jobs. Believe it or not, it does happen, and we can look at one of the world’s largest retailers to see that.
Thanks in part to more robots in its fulfillment centers, Amazon has been able to drive down shipping costs and pass those savings on to customers. Cheaper shipping made more people use Amazon, and the company hired more workers to meet this increased demand.
Division of labor
So what do the robots do, and what do the people do?
Tasks involving fine motor skills, judgment or unpredictability are handled by people. They stock warehouse shelves with items that come off delivery trucks. A robot could do this, except that to maximize shelf space, employees are instructed to stack items according to how they fit on the shelf rather than grouping them by type.
Robots can only operate in a controlled environment, performing regular and predictable tasks. They’ve largely taken over heavy lifting, including moving pallets between shelves—good news for warehouse workers’ backs—as well as shuttling goods from one end of a warehouse to another.
Under current technology, the expense of building robots able to stock shelves based on available space is more costly and less logical than hiring people to do it.
Similarly, for outgoing orders, robots do the lifting and transportation, but not the selecting or packing. A robot brings an entire shelf of goods to an employee’s workstation, where the employee selects the correct item and puts it on a conveyor belt for another employee to package. By this time, the shelf-carrying robot is already returning the first shelf and retrieving another.
Since loading trucks also requires spatial judgment and can be unpredictable—space must be maximized here even more than on shelves—people take care of this too.
More robots mean more humans, for now
Ever since acquiring Boston-based robotics company Kiva Systems in March 2012—at a price tag of $775 million—Amazon has been ramping up its use of robots and is continuing to pour funds into automation research, both for robots and delivery drones.
In 2016 the company grew its robot workforce by 50 percent, from 30,000 to 45,000. Far from laying off 15,000 people, though, Amazon increased human employment by around 50 percent in the same period of time.
Even better, the company’s Q4 2016 earnings report included the announcement that it plans to create more than 100,000 new full-time, full-benefit jobs in the US over the next 18 months. New jobs will be based across the country and will include various types of experience, education, and skill levels.
So how tight is the link between robots and increased productivity? Would there be even more jobs if people were doing the robots’ work?
Well, picture an employee walking (or even running) around a massive warehouse, locating the right shelf, climbing a ladder to reach the item he’s looking for, grabbing it, climbing back down the ladder (carefully, of course), and walking back to his work station to package it for shipping. Now multiply the time that whole process took by the hundreds of thousands of packages shipped from Amazon warehouses each day.
Lots more time. Lots less speed. Fewer packages shipped. Higher costs. Lower earnings. No growth.
Though it may not last forever, right now Amazon’s robot-to-human balance is clearly in employees’ favor. Automation can take jobs away, but sometimes it can create them too.
Image Credit: Tabletmonkeys/YouTube Continue reading