Lifeboat Foundation

After years of hype, many people feel AI has failed to deliver, says Tim Cross.

Technology Quarterly Jun 11th 2020 edition

Our newest water-seeking rover just booked a ride to the Moon’s South Pole.

Pittsburgh-based Astrobotic has been selected to deliver VIPER to the Moon in 2023 in preparation for future #Artemis missions to bring humanity to the lunar surface: https://go.nasa.gov/2YsxZFw

A memristor¹ has been proposed as an artificial synapse for emerging neuromorphic computing applications^2,3. To train a neural network in memristor arrays, changes in weight values in the form of device conductance should be distinct and uniform³. An electrochemical metallization (ECM) memory^4,5, typically based on silicon (Si), has demonstrated a good analogue switching capability^6,7 owing to the high mobility of metal ions in the Si switching medium⁸. However, the large stochasticity of the ion movement results in switching variability. Here we demonstrate a Si memristor with alloyed conduction channels that shows a stable and controllable device operation, which enables the large-scale implementation of crossbar arrays. The conduction channel is formed by conventional silver (Ag) as a primary mobile metal alloyed with silicidable copper (Cu) that stabilizes switching. In an optimal alloying ratio, Cu effectively regulates the Ag movement, which contributes to a substantial improvement in the spatial/temporal switching uniformity, a stable data retention over a large conductance range and a substantially enhanced programmed symmetry in analogue conductance states. This alloyed memristor allows the fabrication of large-scale crossbar arrays that feature a high device yield and accurate analogue programming capability. Thus, our discovery of an alloyed memristor is a key step paving the way beyond von Neumann computing.

Here are a few details that have remained more or less intact: J, when he was alive, had dark brown hair, large eyes, olive skin. He loved Proust and Fitzgerald, Marsalis and Bechet. Said things like “this too shall pass,” wrote them on rogue scraps of paper, and hid them in the bottoms of our backpacks.

Then, one morning, he was gone.

I tried to remember all of the details that constituted him, but my memory would not allow it. Instead, I saw a body vaulting through the sky, a bike clumsily skidding below, and I was furious with myself for not being able to imagine something else or, at the very least, remember more. Me, the keeper of a thousand notebooks. The guardian of a thousand pens. But The Cloud, my trusty companion, seemed to have stored all of him. And when grief took me, it preempted my needs like any good lover would; showed me parts of myself I didn’t know were there and parts of others I had almost completely forgotten—and because the bereaved and the writer in me were rattled in their obsessive need to remember, they both gave themselves over to it entirely.

Rice University researchers have demonstrated methods for both designing innovative data-centric computing hardware and co-designing hardware with machine-learning algorithms that together can improve energy efficiency by as much as two orders of magnitude.

Advances in machine learning, the form of artificial intelligence behind self-driving cars and many other high-tech applications, have ushered in a new era of computing—the data-centric era—and are forcing engineers to rethink aspects of computing architecture that have gone mostly unchallenged for 75 years.

“The problem is that for large-scale deep neural networks, which are state-of-the-art for machine learning today, more than 90% of the electricity needed to run the entire system is consumed in moving data between the memory and processor,” said Yingyan Lin, an assistant professor of electrical and computer engineering.

Last week, I wrote about the diversity problem in AI and why we need to fix it. I asked you to tell us about your experiences as a Black person in AI or share the names of Black colleagues you admire. Thank you to everyone who responded. It was heart-warming to hear from so many of you.

Many of you shared your frustration with the lack of mentors who understand your challenges, the alienation of being the only Black face at professional meetings, and the struggle to overcome economic and social inequalities. Black women, especially, wrote about the difficulties of building a career in AI. Some of you described your efforts to support Black people in science and technology and provide tech resources to underserved communities. Thank you for sharing with us your dreams and also your disappointments.

We will feature some of your stories in our Working AI blog series. Please stay tuned.

A research team from ETH Zurich and the Max Planck Institute for Solar System Research in Göttingen counted over 136,000 rockfalls on the moon caused by asteroid impacts. Even billions of years old landscapes are still changing.

In October 2015, a spectacular rockfall occurred in the Swiss Alps: in the late morning hours, a large, snow-covered block with a volume of more than 1500 cubic meters suddenly detached from the summit of Mel de la Niva. It fell apart on its way downslope, but a number of boulders continued their journey into the valley. One of the large boulders came to a halt at the foot of the summit next to a mountain hut, after traveling more than 1.4 kilometers and cutting through woods and meadows.

On the moon, time and again boulders and blocks of rock travel downslope, leaving behind impressive tracks, a phenomenon that has been observed since the first unmanned flights to the moon in the 1960s. During the Apollo missions, astronauts examined a few such tracks on site and returned displaced rock block samples to Earth. However, until a few years ago, it remained difficult to gain an overview of how widespread such rock movements are and where exactly they occur.

We have heard of the latest advancements in the field of deep learning due to the usage of different neural networks. Most of these achievements are simply astonishing and I find myself amazed after reading every new article on the advancements in this field almost every week. At the most basic level, all such neural networks are made up of artificial neurons that try to mimic the working of biological neurons. I had a curiosity about understanding how these artificial neurons compare to the structure of biological neurons in our brains and if possibly this could lead to a way to improve neural networks further. So if you are curious about this topic too, then let’s embark on a short 5-minute journey to understand this topic in detail…

Researchers at the Nanoscience Center and at the Faculty of Information Technology at the University of Jyväskylä in Finland have demonstrated that new distance-based machine learning methods developed at the University of Jyväskylä are capable of predicting structures and atomic dynamics of nanoparticles reliably. The new methods are significantly faster than traditional simulation methods used for nanoparticle research and will facilitate more efficient explorations of particle-particle reactions and particles’ functionality in their environment. The study was published in a Special Issue devoted to machine learning in the Journal of Physical Chemistry on May 15, 2020.

The new methods were applied to ligand-stabilized metal nanoparticles, which have been long studied at the Nanoscience Center at the University of Jyväskylä. Last year, the researchers published a method that is able to successfully predict binding sites of the stabilizing ligand molecules on the nanoparticle surface. Now, a new tool was created that can reliably predict potential energy based on the atomic structure of the particle, without the need to use numerically heavy electronic structure computations. The tool facilitates Monte Carlo simulations of the atom dynamics of the particles at elevated temperatures.

Potential energy of a system is a fundamental quantity in computational nanoscience, since it allows for quantitative evaluations of system’s stability, rates of chemical reactions and strengths of interatomic bonds. Ligand-stabilized metal nanoparticles have many types of interatomic bonds of varying chemical strength, and traditionally the energy evaluations have been done by using the so-called density functional theory (DFT) that often results in numerically heavy computations requiring the use of supercomputers. This has precluded efficient simulations to understand nanoparticles’ functionalities, e.g., as catalysts, or interactions with biological objects such as proteins, viruses, or DNA. Machine learning methods, once trained to model the systems reliably, can speed up the simulations by several orders of magnitude.