A secretive startup has unveiled its new grocery-getter, one of many automakers that are leaving people out of the equation.
Read more
Category: information science – Page 284
A new model can detect abnormalities in x-rays better than radiologists—in some parts of the body, anyway.
The results: Stanford researchers trained a convolutional neural network on a data set of 40,895 images from 14,982 studies. The paper documents how the algorithm detected abnormalities (like fractures, or bone degeneration) better than radiologists in finger and wrist radiographs. However, radiologists were still better at spotting issues in elbows, forearms, hands, upper arms, and shoulders.
The background: Radiologists keep getting put up against AI, and they usually don’t fare even as well as this. Geoffrey Hinton, a prominent AI researcher, told the New Yorker that advances in AI mean that medical schools “should stop training radiologists now.”
Read more
Neuroscientist Brie Linkenhoker believes that leaders must be better prepared for future strategic challenges by continually broadening their worldviews.
As the director of Worldview Stanford, Brie and her team produce multimedia content and immersive learning experiences to make academic research and insights accessible and useable by curious leaders. These future-focused topics are designed to help curious leaders understand the forces shaping the future.
Worldview Stanford has tackled such interdisciplinary topics as the power of minds, the science of decision-making, environmental risk and resilience, and trust and power in the age of big data.
Read more
Standard utopias and standard dystopias are each perfect after their own particular fashion. We live somewhere queasier—a world in which technology is developing in ways that make it increasingly hard to distinguish human beings from artificial things. The world that the Internet and social media have created is less a system than an ecology, a proliferation of unexpected niches, and entities created and adapted to exploit them in deceptive ways. Vast commercial architectures are being colonized by quasi-autonomous parasites. Scammers have built algorithms to write fake books from scratch to sell on Amazon, compiling and modifying text from other books and online sources such as Wikipedia, to fool buyers or to take advantage of loopholes in Amazon’s compensation structure. Much of the world’s financial system is made out of bots—automated systems designed to continually probe markets for fleeting arbitrage opportunities. Less sophisticated programs plague online commerce systems such as eBay and Amazon, occasionally with extraordinary consequences, as when two warring bots bid the price of a biology book up to $23,698,655.93 (plus $3.99 shipping).
In other words, we live in Philip K. Dick’s future, not George Orwell’s or Aldous Huxley’s. Dick was no better a prophet of technology than any science fiction writer, and was arguably worse than most. His imagined worlds jam together odd bits of fifties’ and sixties’ California with rocket ships, drugs, and social speculation. Dick usually wrote in a hurry and for money, and sometimes under the influence of drugs or a recent and urgent personal religious revelation.
Still, what he captured with genius was the ontological unease of a world in which the human and the abhuman, the real and the fake, blur together. As Dick described his work (in the opening essay to his 1985 collection, I Hope I Shall Arrive Soon):
Read more
From afar, it looks like a steampunk chandelier. An intricate collection of tubes and wires that culminate in a small steel cylinder at the bottom. It is, in fact, one of the most sophisticated quantum computers ever built. The processor inside has 50 quantum bits, or qubits, that process tasks in a (potentially) revolutionary way. Normally, information is created and stored as a series of ones and zeroes. Qubits can represent both values at the same time (known as superposition), which means a quantum computer can theoretically test the two simultaneously. Add more qubits and this hard-to-believe computational power increases.
Last November, IBM unveiled the world’s first 50-qubit quantum computer. It lives in a laboratory, inside a giant white case, with pumps to keep it cool and some traditional computers to manage the tasks or algorithms being initiated. At CES this year, the company brought the innards — the wires and tubes required to send signals to the chip and keep the system cool — so reporters and attendees could better understand how it works. The biggest challenge, IBM Research Vice President Jeffrey Welser told me, is isolating the chip from unwanted “noise.” This includes electrical, magnetic and thermal noise — just the temperature of the room renders the whole machine useless.
Read more
Today at CES, Samsung unveiled DRVLINE, a hardware and software platform that will allow car makers to create customized, technologically advanced autonomous vehicles. Many platforms are an all-or-nothing solution, which forces users to adopt the entire package en masse, without any sort of customization. DRVLINE, however, allows vendors to swap and customize individual components, building the vehicle to their specifications, as well as allowing for rapidly evolving technology.
“Building an autonomous platform requires close collaboration across industry, as one company cannot deliver on this enormous opportunity alone,” said Young Sohn, the president and chief strategy officer of Samsung. “The challenge is simply too big and too complex. Through the DRVLINE platform, we’re inviting the best and brightest from the automotive industry to join us, and help shape the future of the car of tomorrow, today.”
The first DRVLINE initiative will be a camera that features lane departure warnings, adaptive cruise control, collision warning and algorithms that can deliver warnings about pedestrians. The system will start shipping this year.
Read more
Wirelessly powered, biomimetic spybots…
A new contest seeks flight systems inspired by Mother Nature and powered by directed-energy beams.
Tired: multi-rotor copters and fixed-wing drones. Wired: flying robots that move like living animals, are crafted of next-generation materials, and draw their power not from batteries but energy beamed from nearby aircraft.
Leave A.I. Alone
Posted in information science, law, robotics/AI
December was a big month for advocates of regulating artificial intelligence. First, a bipartisan group of senators and representatives introduced the Future of A.I. Act, the first federal bill solely focused on A.It would create an advisory committee to make recommendations about A.I. — on topics including the technology’s effect on the American work force and strategies to protect the privacy rights of those it impacts. Then the New York City Council approved a first-of-its-kind bill that once signed into law will create a task force to examine its own use of automated decision systems, with the ultimate goal of making its use of algorithms fairer and more transparent.
Sure, the technology poses risks. But the current approach to regulating it is a mistake.
Read more
China has unveiled three-year plans to increase the country’s economic competitiveness by developing “key technologies” in nine industrial sectors, from robotics to railways.
Other areas include smart cars, robotics, advanced shipbuilding and maritime equipment, modern agricultural machinery, advanced medical devices and drugs, new materials, smart manufacturing and machine tools.
The aim is “to make China a powerful manufacturing country” and upgrade the nation’s industrial power through “the internet, big data and artificial intelligence”, the commission said.
To achieve that goal, the agency has laid out specific targets to develop key technologies and guide research and the flow of funds in each sector.
After less than eight months of development, the algorithms are helping intel analysts exploit drone video over the battlefield.
Earlier this month at an undisclosed location in the Middle East, computers using special algorithms helped intelligence analysts identify objects in a video feed from a small ScanEagle drone over the battlefield.
A few days into the trials, the computer identified objects — people, cars, types of building — correctly about 60 percent of the time. Just over a week on the job — and a handful of on-the-fly software updates later — the machine’s accuracy improved to around 80 percent. Next month, when its creators send the technology back to war with more software and hardware updates, they believe it will become even more accurate.
Read more