Robopocalypse Report #11

This time we have a droneless report!  Woo!  A first!  So we'll start with the next most photogenic of the Robopocalypse: the self driving car.

CNET runs an article about how the self driving car is driving innovation in computer objection recognition.

Future producer of self driving cars (well, almost there for highway driving now) Tesla had its Model S hacked.  This required physical access to the car unlike the other car hacks recently.  However, this is my greatest concern with the self driving car.  I would love to use it as a way to pick up my kids and drop them off for lessons and whatnot.  I cannot, not just because of the foreseeable requirement to have an adult 'driving' because I can all too easily see a hacker-kidnapper taking control of the car to deliver the kids inside wherever rather than to where they were supposed to go.

Would be self driving taxi car company Uber is apparently bleeding out money rather quickly and if reports are true, you have to wonder if they are going to survive long enough to field their fleet for the robopocalypse.

Moving away from the self driving car, MIT is working on improving the teleoperation of robots.  In some ways this is not new technology: it was even used for the Johnny 5 robot in Short Circuit 2.  However, I can only imagine with modern computers its far, far better.

There have been increasing calls to ban autonomous weapons.  Everyone is invoking the Terminator for this.  Others have started calling them weapons of mass destruction in their calls for a ban.  I had an essay I was working on about the overuse of the term 'mass extinction' on an old computer, but the computer was lost and I didn't have a backup.  It seems WMD is hitting that level, too.  OTOH, autonomous weapons?  What are we talking here?  The mythical forever flying machines with solar powered lasers?  Or a mine?  Anything which requires human permission before firing?  Or a 5th generation atlas with an AK?  The definition above is really, really broad and will never make it into reality based on what is being proposed here.  

The Japanese have their robo hotel with the really odd kiosks for checking in.  The brits have taken their own version of the hotel of the robopocalypse.  Here's a review of the "Hub by Premier Inn."

And to round out the Robopocalypse Report #11, IBM has purchased Merge Healthcare, a medical imagery company, for use with its Watson AI system for a cool BILLION dollars.  Watson is either going to make IBM incredibly wealthy or crash the company.  Is this IBM's 747 moment?

IBM Makes Progress Towards Quantum Computing

IBM scientists have unveiled two crucial advances toward the creation of a practical quantum computer: an effective way to detect and correct quantum errors, and the design of a silicon chip that can scale up to house a large number of entangled quantum bits.

link.

Tip Toeing to the Robopocaylpse With IBM

A million spiking-neuron integrated circuit with a scalable communication network and interface

Authors:

Merolla et al

Abstract:

Inspired by the brain’s structure, we have developed an efficient, scalable, and flexible non–von Neumann architecture that leverages contemporary silicon technology. To demonstrate, we built a 5.4-billion-transistor chip with 4096 neurosynaptic cores interconnected via an intrachip network that integrates 1 million programmable spiking neurons and 256 million configurable synapses. Chips can be tiled in two dimensions via an interchip communication interface, seamlessly scaling the architecture to a cortexlike sheet of arbitrary size. The architecture is well suited to many applications that use complex neural networks in real time, for example, multiobject detection and classification. With 400-pixel-by-240-pixel video input at 30 frames per second, the chip consumes 63 milliwatts.

pop sci write up.

IBM's Watson Developer Challenge to Bring Cognitive Apps to Main Stream

Of IBM’s efforts to transform itself over the last few years, few have captured the public’s imagination like the company’s Watson supercomputer.

Since its first appearance on “Jeopardy” in 2011, many have pointed to Watson as an indicator of where we’re headed with technology like Apple’s Siri.

If fed enough data, Watson can become an expert on a seemingly unlimited number of topics. Using probabilities, it’s able to work out how different concepts relate to each other and find solutions to problems.

The company has even said that the next generation will be able to reason and debate. Now it’s pushing to get developers to build that power into the apps you use everyday on your smartphone.

IBM first opened up Watson for development as a cloud platform back in November. To encourage its adoption, the company today announced the Watson Developer Challenge, a three-month-long competition in which developers can submit ideas and prototypes for so-called cognitive apps, or “cogs,” that could benefit from Watson’s natural language processing capabilities.

At the end of the three-month challenge, IBM will help three winners turn their ideas into commercial apps by providing them with seed funding, design consulting, and support from IBM Interactive Experience, its new global consulting practice.

IBM has already received pitches for apps utilising Watson from over 1,500 individuals and organisations, according to a press release sent out earlier today.

link.

Robopocalypse: IBM Makes 'Creative' Computer AI

Can computers be creative? That’s a question likely to generate controversial answers. It also raises and some important issues too, like how to define creativity.

Seemingly unafraid of the controversy, IBM has darted into the fray by answering this poser with with a resounding ‘yes’. Computers can be creative, they say, and to prove it they have built a computational creativity machine that produces results that a knowledgeable human would consider novel, useful and even valuable—the hallmarks of genuine creativity.

IBM’s chosen field for this endeavour is cooking. The company’s creativity machine produces recipes based on chosen ingredients or cooking styles. And they’ve asked professional chefs to evaluate the results and say the feedback is promising.

link.

NERSC's New System: Announcing Carver

Carver: IBM iDataPlex

NERSC's next medium-sized scientific computing system will be an IBM iDataPlex Linux cluster. The IBM system, selected in a competitive procurement, provides excellent performance, good energy efficiency per flop, and a familiar environment for mid-range parallel applications. It is intended to replace Bassi and Jacquard.

The system will be named after the American scientist George Washington Carver.

410 nodes with dual quad core Nehalems. IB interconnect. More than 10 TB of memory. 1.2 PB of disk. 34.2 TFlops peak. Link in title, as always.

10 Million X Finer Resolution MRI

IBM Research scientists, in collaboration with the Center for Probing the Nanoscale at Stanford University, have demonstrated magnetic resonance imaging (MRI) with volume resolution 100 million times finer than conventional MRI.

This result, published today in the Proceedings of the National Academy of Sciences (PNAS), signals a significant step forward in tools for molecular biology and nanotechnology by offering the ability to study complex 3D structures at the nanoscale.

By extending MRI to such fine resolution, the scientists have created a microscope that, with further development, may ultimately be powerful enough to unravel the structure and interactions of proteins, paving the way for new advances in personalized healthcare and targeted medicine. This achievement stands to impact the study of materials from proteins to integrated circuits for which a detailed understanding of atomic structure is essential.

Today has been a kewl day for scitech announcements.

IBM isn't out from under the death glare as work though.

Moscow State University Gets Blue Gene/P

Moscow State University (MSU) and IBM (NYSE: IBM) today announced details of an agreement to install a Blue Gene/P supercomputer at the Department of Computational Mathematics and Cybernetics. The new supercomputer, the first installation of the world famous Blue Gene computer in Russia, will be used for fundamental research in nanotechnology, new materials and life sciences.

"As Russia's leading academic institution, we are very proud that Moscow State University should join the ranks of the world's leading research organisations to tackle some of the most complex and computer intensive problems known to man-kind -- from astrophysics, to molecular modelling," said Viktor Sadovnichiy, Rector of Moscow State University. "This agreement with IBM heralds a new era of supercomputing in Russia."

Moscow State University has agreed to buy two racks of the Blue Gene/P system containing 8.192 densely packed microprocessors. The resulting supercomputer is expected to run at 27.8 trillion operations per second (Tflop/s), 2,600 times faster than today's fastest home PC. Based on the current Top500.org list, the MSU system would today be amongst the top 50 most powerful supercomputers in the world.

Good for them. Keep in mind that's a *PEAK* speed of 27.8 teraflops. It's probably a sustained speed of around 10 TFlops, maybe 18 Tflops, on benchmarks and even less - a lot less - on real world codes. Even so, the BG/P's are good systems. It's just a consequence of modern computer tech (and some coding practices) that make this true.