Reaction to Persistent Theft at Wal-Mart

 Have now seen several examples of this,  reaching epidemic levels

Will persistent theft compel Walmart to raise prices or close stores?

32 EXPERT COMMENTS  by George Anderson   in RetailWire

Walmart CEO Doug McMillon yesterday told CNBC that his company has a growing theft problem

Mr. McMillon said that theft has reached levels that are “higher than what’s historically been” the case for the retailer. “If that’s not corrected over time, prices will be higher, and/or stores will close,” he said.

Walmart’s CEO said the company has put safety and security measures in place to address theft. He said that “local law enforcement being staffed and being a good partner” is part of the response needed to curtail the frequency with which goods are illegally walked out of stores.

In response to a question about laws that make it more difficult to prosecute shoplifters, Mr. McMillon said that laws and practices vary by location and that the retailer would benefit from greater “policy consistency and clarity” so that it would be able to make  capital investments where they make the most sense in thwarting thieves.

Walmart’s McMillon also offered a nuanced explanation of how inflation is affecting consumers in the CNBC interview.

“It’s probably helpful to break that [inflation] down into different categories. Fresh food is more volatile, it fluctuates more, so today’s beef prices are down and chicken prices are still high. Produce prices are relatively low relative to what they were before,” he said. “Dry grocery, processed foods and consumables are where the inflation is most stubborn. Double digit inflation rates have been around for a while and it looks to us like they’re going to be with us for a while. So that’s a primary area of focus for us — trying to come up with creative ways with our suppliers to get prices down. Customers choose private brands in those environments more often.”

Inflation, he said, was less of a problem in general merchandise categories such as apparel, sporting goods and toys where “prices have come down more aggressively.” Mr. McMillon said.

“We’re still inflated but we’re not inflated nearly as much as we are in the other categories,” he said. “So when you think about what’s happening, families are choosing when they buy their big basket of things that they need all the time to look for value. Many of them are coming to us for those big food and consumables baskets, and then they’re being selective on general merchandise.”  ... ' 

Connecting the Joint Force as One

Northrup Grumman:   Connecting the Joint Force as One

Future mission success will lay in highly-effective Joint All-Domain Command and Control (JADC2)

‘Joining Forces’ has always been a critical factor in creating extra strength

In the future, more than ever before, mission advantage will go to military forces who can coordinate and act together across all domains. Observing, orienting, deciding and acting together, faster and with greater effect than their adversaries is the next-gen technological edge for Joint Forces. And, whether Joint Forces are across United States military branches including the U.S. Air Force, U.S. Army, U.S. Navy, U.S. Marines, U.S. Space Force, or in partnership with allies―future mission success will lay in highly-effective Joint All-Domain Command and Control (JADC2).

Ars Technicast Special Edition: The Internet of Military Things

For decades, the US Department of Defense has been trying to make the tools we use to fight battles more like the tools we use in peacetime with more connectivity, more data, and more smarts. The difference this time around, is that technology has caught up to the point where what we can do in real life almost mirrors what we can do in movies. To learn more, listen to Scott Stapp, Chief Technology Officer, Northrop Grumman on Ars Technica’s podcast.

(Podcast) 

Robot Design May Revolutionize How We Build Things in Space

Obviously, but what developments should we wait for?  Specify?  Need?

Robot Design May Revolutionize How We Build Things in Space

SciTechDaily, December 1, 2022

Researchers at the U.K.'s University of Lincoln (UoL) designed an end-over-end walking (E-Walker) robot for space construction projects. The E-Walker robot features seven degrees-of-freedom motion capabilities. UoL's Manu Nair and colleagues conducted an exercise in which they tested the E-Walker on the in-orbit assembly of a 25-meter (82-foot) Large Aperture Space Telescope. Nair said, "The proposed innovative E-Walker design proves to be versatile and an ideal candidate for future in-orbit missions. The E-Walker would be able to extend the life cycle of a mission by carrying out routine maintenance and servicing missions post assembly, in space." .... ' 

How China's Police Used Phones, Faces to Track Protesters

 China Policing

How China's Police Used Phones, Faces to Track Protesters

The New York Times

Paul Mozur; Claire Fu; Amy Chang Chien, December 2, 2022

China's police used an advanced surveillance system to track protesters who rallied against the government's pandemic policies this past week. The system enables authorities to target, detain, and intimidate protest organizers and vocal dissidents. Its tools include millions of cameras, facial recognition software programmed to identify local citizens, phone monitors, and data- and image-crunching applications. The phone trackers connect to and record data on the phones of passersby for police to review. Many protesters said they now avoid using virtual private networks or other foreign apps like Telegram and Signal out of fear their phones' software might be more closely monitored.  ... '

RobotTaxis to Thrive in Vegas

 Where Else?  Good place for live tests, proof of use in complex use, Uber complexity ...  On to LA they say ... 

Uber and Motional's robotaxis arrive in Las Vegas

It marks the first time public riders and access the autonomous vehicles using Uber.

Steve Dent|@stevetdent|December 7, 2022 6:00 AM

Uber has launched public robotaxi rides in Las Vegas using Motional's Hyundai Ioniq 5 autonomous EVs with the aim of offering a full driverless service to the public in 2023. It will eventually expand to Los Angeles, where the two companies have been testing autonomous Uber Eats deliveries since May 2022. It's all part of a 10-year agreement between Uber and Motional to offer autonomous ride-hailing and deliveries. 

The taxi rides will be monitored by safety drivers, with the goal of launching a fully driverless service to the public in 2023. "Today, Motional becomes the first AV company to conduct all-electric autonomous rides on the Uber network for public passengers," said Motional VP Akshay Jaising. (Uber offered "autonomous" taxi rides using its own self-driving tech back in 2016, but riders were accompanied by engineers ready to take the wheel.) 

Lyft also partners with Motional, a joint venture between Hyundai and Aptiv. It beat Uber to the punch by launching Motional rides in Las Vegas in August this year with safety drivers on board. Lyft has also said it will ditch those drivers and offer true autonomous service by next year. 

Don't be surprised if that timeline changes, though. Apart from Motional, only Alphabet division Waymo and GM's Cruise are offering true driverless services at a reasonably large scale. The Waymo One service is operating in Phoenix and San Francisco, while Cruise rides are currently limited to San Francisco. Both operate only in specific areas of cities (which can be mapped out in great detail) and some vehicles still use safety riders.

Uber co-founder Travis Kalanick once said that the his company had to be "tied for first, at least" in the race to offer true driverless rides. Since both Uber and Lyft rely on Motional for self-driving tech, that's exactly what might happen. ... ' 

AI Helps Generate Building Blocks for New Drugs

Generating New Drugs

AI Helps Biotech Labs Generate the Building Blocks for New Drugs

By Adrianna Nine on December 6, 2022 at 7:56 am Comments

Myotis daubentonii. (Credit: Generate Biomedicines)

Proteins are an essential part of life. Not only do they function as the “building blocks” for living organisms, but they also perform nearly every cellular task, from waste management to tissue repair. It tracks, then, that pharmaceuticals often contain or “target” proteins in an attempt to change or eliminate symptoms or disease within the body. There’s just one little problem: The only proteins we can use to create drugs are the ones we know.

But what if we could add new proteins to the resource pile? Two biotech labs have begun using artificial intelligence (AI) to generate novel proteins with the goal of incorporating them into medicine. Chroma, a program by Boston-based Generate Biomedicines, and RoseTTAFold Diffusion, a program out of the University of Washington’s Baker Lab, are often compared with trendy AI art generators thanks to their ability to create new concepts seemingly out of thin air. In fact, Chroma has even been referred to as “the DALL-E 2 of biology.”

AI protein generators, a type of denoising diffusion probabilistic model (DDPM), are trained to pull samples from complex datasets. They then use those samples—plus added noise—to construct a product that matches a given prompt. Although art generators are the best-known DDPMs, AI isn’t only good for producing images on demand; they’ve also been found to be an ideal method of generating diverse protein models. In a preprint on bioRxiv, Baker Labs scientists describe how RoseTTAFold can devise protein structures with virtually any desired size, shape, or function, as well as with any required constraints. ... ' 

Setting the Internet of Things Free -- of Batteries

Yes, please.

Setting the Internet of Things Free -- of Batteries

By R. Colin Johnson, Commissioned by CACM Staff, November 8, 2022

EnOcean’s wall-mounted pushbuttons harvest the energy from the finger pushing them to generate an RF signal that turns on lights, overhead fans, or any other device that would ordinarily require connection to electrical wiring.

Energy harvesting reaps energy that would otherwise go to waste—from sources ranging from solar rays to vibrations to ambient gradations in heat, ocean waves/tides, wind, and even the metabolisms of living organisms. These sources can be tuned to produce electricity that is safe, sustainable, and deployable in conditions not suited to the use of batteries.

Incorporating such energy harvesting into the Internet of Things (IoT) extends its reach to nearly any place on the globe—not just those that can access power lines or rechargeable batteries.

"Energy harvesting has the potential to expand the capabilities of the Internet of Things to heights previously unattainable without the technology," said ABI Research analyst Taylor Jensen, who is just putting the finishing touches on a study on energy harvesting for IoT applications.

Applications for IoT devices are expanding dramatically, targeting industries including automotive, aerospace, military, transportation, consumer electronics, industrial, buildings, home appliances, environmental monitoring, home automation, healthcare, and other increasingly "green" sectors. Ubiquitous communication standards such as Bluetooth, Wi-Fi, and cellular technologies have opened the door to IoT devices that monitor and manage widespread networks of sensors and actuators. With the addition of energy harvesting to power off-grid IoT devices, there are few corners of the Earth left that are out of reach.

Energy harvesting has matured through the use of a variety of environmental sources, including:

• Light (via photovoltaic cells).

• Wind (which drives turbines that generate electricity).

• Vibration (resonators drive transducers to generate electricity).

• Thermal (temperature differences converted to electrical energy by dipole oscillators)

• Radio Waves (electricity harvested from oscillating electromagnetic fields with antennas and diodes).

Today, IoT devices often depend on just one of the energy harvesting methods above; for example, solar cells work fine on sunny days, but not so well on cloudy days or at night. To compensate, the solar cells charge a battery during the day, then switch to battery back-up systems to provide power at night.

Yet batteries are not the long-term power solution for the IoT, according to Pat Pannuto, an assistant professor in the Department of Computer Science and Engineering at the University of California, San Diego.

"Every device with a battery is a device with a lifetime, and a short one at that," said Pannuto. "Today, we already average 10 or more connected devices per person in many parts of the world. Now imagine the transition from today's Internet of Things to tomorrow's Internet of Everything, a future with as many as a trillion connected devices; that is, hundreds of devices per person on the planet. We cannot become a world of battery-tenders; self-sufficient devices are fundamental to scale."

One answer available today has been demonstrated by Ji Li, a senior data and applied science manager at Microsoft. Li says 24/7/365 battery-free always-on IoT devices merely need to include multiple energy harvesters. Rather than using a single-source harvester that stores energy in a battery for times of darkness, no wind, etc., Li has demonstrated a method that combines several energy sources that together can provide the power requirements of an IoT device all day, every day, without the need for batteries.

According to Li, battery-free IoT devices require the use of machine learning (ML) to intelligently switch between multiple energy sources customized to provide the most efficient and most stable power supply for a particular application. For instance, managing the switch among thermal, kinetic, and photovoltaic energy sources with algorithms that compensate for the intermittent nature of each individual energy source allows ML to inform task-scheduler software that automatically switches among power sources.

"Batteries are the limiting factor in the lifetime of today's electronic devices. To extend service lifetime, devices must be able to obtain energy from their deployment context," said Pannuto. "That means scavenging energy from the nearby environment. As we push into more diverse environments, this also means scavenging from new and novel sources. In the future, intelligent instrumented systems will run on energy found wherever they are deployed. For instance, sensors embedded in concrete can harvest from the corrosion of the rebar. Sensors buried underground can harvest from electrogenic bacteria in the soil."  ... ' 

Free-Space Optics System Could Boost Space Comms

Free Space Optics

Free-Space Optics System Could Boost Space Comms

IEEE Spectrum. John Boyd.December 1, 2022

Researchers at Australia's International Center for Radio Astronomy Research (ICRAR) established a laser-guided free-space optical (FSO) link with the potential to enhance communications for space missions. The researchers maintained the FSO link despite stormy atmosphere between an optical ground terminal and a retroreflector on an airborne drone. ICRAR's Shane Walsh explained the framework supports "an uninterrupted 100-gigabits-per-second optical-data link. We do this by tracking the drone at angular rates up to 1.5 degrees a second—the equivalent of tracking a satellite in low Earth orbit [LEO]." Walsh said this clears a path for realizing terabit-per-second communications between LEO satellites and ground stations. ... ' 

Three Key Takeaways from Meta's Galactica AI

Galactica AI,  general idea is good, navigating complex forms of data is useful.  One of our early AI successes did exactly this for key internal knowledge.  Make it easy to easy to use, and importantly make it easy to validate and update in changing contexts.

Three Takeaways from Meta's Galactica AI

By TechTalks, November 23, 2022

We should look at Galactica's initial failure as another scientific experiment...and, every failed experiment brings us one step closer to success.

Meta's most recent large language model (LLM) release, Galactica, was supposed to help scientists navigate tons of published scientific information. Its developers presented it as being able to find citations, summarize academic literature, solve math problems, and perform other tasks that help scientists in research and writing papers.

However, three days after the release, Meta had to shut down the online demo following a deluge of criticism by scientists and tech media about the model's incorrect and biased output. While Galactica was obviously not a success, its short history provides us with some useful lessons about LLMs and the future of AI research.

From TechTalks   View Full Article    

The State of Grocery Retail Around the World

The state of grocery retail around the world

October 6, 2022 | Podcast     By Bill Aull, Dymfke Kuijpers, and Daniel Läubli  in McKinsey

The state of grocery retail around the world ... 

A survey of more than 30,000 consumers and 100 grocery CEOs reveals the challenges—and opportunities—that grocery retailers face today.

Changes to the global grocery retail market are happening fast. Trends that emerged during the pandemic are persisting, reshaping the industry. Consider this: in December 2019, the percentage of e-commerce grocery sales in the US market was slightly under 4 percent; it’s now in the low teens and is expected to reach roughly 25 percent by 2030. The adoption of online grocery is expected to be even higher in parts of Europe.

In this episode of the McKinsey on Consumer and Retail podcast, three McKinsey experts share insights from their latest research on the global grocery retail industry. An edited transcript of their conversation with host Monica Toriello follows. Subscribe to the podcast.

Monica Toriello: Today we’re going to hear about the state of grocery retail all around the world. McKinsey’s Retail Practice has undertaken a huge research effort that encompassed surveys of more than 30,000 consumers and more than 100 grocery CEOs globally about what’s happening in grocery and what the future of the sector might look like. We’ll hear from the regional leaders of this research to tell us about what they’re seeing in Asia, Europe, and North America.

First up is Bill Aull, a partner who leads McKinsey’s office in Charlotte, North Carolina. Bill has expertise in a number of retail and consumer packaged goods (CPG) topics, including merchandising, category management, and strategy. Bill, thanks for being here—and tell us one or two things that you get at every trip to the grocery store. For me, it’s seltzer water and some kind of dried fruit. What is it for you?

Most Popular Insights

Author Talks: What poker pro Annie Duke can teach you about quitting on time

Europe’s EV opportunity—and the charging infrastructure needed to meet it

A devilish duality: How CEOs can square resilience with net-zero promises

Bill Aull: Great to be here. At the grocery store, I always get cream cheese, eggs, and some sort of grainy bread.

Monica Toriello: Next up is Dymfke Kuijpers, a senior partner who leads our Consumer Practice in Asia and our work in consumer tech in the region. Dymfke is a Dutch national who’s been based in Singapore for the past five years. Dymfke, nice to have you with us. And tell us one thing you can’t leave the grocery store without.

Dymfke Kuijpers: Very nice to be here, Monica. I never leave the grocery store without stone berries. They’re the best indication of the supply chain of the grocer.

Monica Toriello: That’s good to know. Last but not least, Daniel Läubli is a senior partner based in Zurich. He leads McKinsey’s work in grocery globally and coleads our Retail Practice in Europe. His areas of expertise include strategy, store operations, and advanced analytics. Welcome to the podcast, Daniel, and tell us what your grocery must-have is.

Daniel Läubli: Monica, great to be here. My grocery must-have is dark Swiss chocolate. I eat at least a pack a day, so whenever I go shopping, I need a lot of that.

How consumer behavior has changed

Monica Toriello: All of us, to some extent, are grocery shoppers; we’re all consumers. So let’s start this conversation with the consumer. Broadly speaking, when consumers shop for groceries these days, they notice that prices are higher. And they’re trying to stretch their budgets, so they’ve become more value conscious, as your grocery reports have said—but you’ve also said they want to eat healthier and they care about sustainability. There’s some tension between those things, because healthier foods and sustainable products tend to be not the cheapest options. Bill, let’s start with North America: Is the picture that I’ve painted an accurate picture of the consumer?  ... ' 

Mechanical Neural Networks

 Technical intro abstract

Mechanical neural networks: Architected materials that learn behaviors

RYAN H. LEE HTTPS://ORCID.ORG/0000-0003-2509-1677, ERWIN A. B. MULDER, AND JONATHAN B. HOPKINS HTTPS://ORCID.ORG/0000-0003-4752-746X Authors Info & Affiliations

SCIENCE ROBOTICS  19 Oct 2022, Vol 7, Issue 71, DOI: 10.1126/scirobotics.abq7278

Abstract

Aside from some living tissues, few materials can autonomously learn to exhibit desired behaviors as a consequence of prolonged exposure to unanticipated ambient loading scenarios. Still fewer materials can continue to exhibit previously learned behaviors in the midst of changing conditions (e.g., rising levels of internal damage, varying fixturing scenarios, and fluctuating external loads) while also acquiring new behaviors best suited for the situation at hand. Here, we describe a class of architected materials, called mechanical neural networks (MNNs), that achieve such learning capabilities by tuning the stiffness of their constituent beams similar to how artificial neural networks (ANNs) tune their weights. An example lattice was fabricated to demonstrate its ability to learn multiple mechanical behaviors simultaneously, and a study was conducted to determine the effect of lattice size, packing configuration, algorithm type, behavior number, and linear-versus-nonlinear stiffness tunability on MNN learning as proposed. Thus, this work lays the foundation for artificial-intelligent (AI) materials that can learn behaviors and properties.  .. ' 

Top Supercomputers

High performance computing.  (HPC) 

Frontier Keeps Top Supercomputer Spot

By HPCwire, November 16, 2022

The Frontier supercomputer.

Top500 coauthor Jack Dongarra acknowledged that not all the systems on the Top500 are being used for scientific computing, adding that one way to get a better sense of the real HPC machines is to look at the top 100 or 200 systems on the list.

The Frontier system at the U.S. Department of Energy's Oak Ridge National Laboratory remained the world's top supercomputer on the 60th Top500 list, clocking 1.102 exaflops and scoring highest on the HPL-MxP benchmark with 7.9 exaflops.

The third-highest-ranking supercomputer, Finland’s Lumi, had the second-highest HPL-MxP score with 2.2 exaflops, while Japan's second-ranked Fugaku supercomputer earned the third-highest HPL-MxP position with 2.0 exaflops.

Fugaku held its lead position on the High Performance Conjugate Gradients benchmark with 16 Linpack petaflops, while Frontier scored second with 14 petaflops.

Lumi grew its computational footprint, raising its score from 151.9 to 309.1 petaflops.

From HPCwire   View Full Article       

Sutherland on the Future of Augmented Reality Displays

In fact, at the big enterprise, we tested some of Sutherland's findings in real life retail.   But found it to mostly be too 'busy' for practical use.  Still worth further examination in a new era of adaptive interaction.  What would make it pop for the right cases? 

The metaverse offers challenges and possibilities for the future of the retail industry  by Omar H. Fares, The Conversation

In 1968, American computer scientist Ivan Sutherland predicted the future of augmented and virtual reality with his concept of the "Ultimate Display. The Ultimate Display relied on the kinetic depth effect to create two dimensional images that moved with its users, giving the illusion of a three-dimensional display.

While the concept of virtual reality only focuses on the creation of three-dimensional environments, the metaverse—a term coined by Neal Stephenson in his 1992 book Snow Crash—is a much broader concept that surpasses this.

While no official definition of the metaverse truly exists, science and technology reporter Matthew Sparkes provides a decent one. He defines the metaverse as "a shared online space that incorporates 3D graphics, either on a screen or in virtual reality."

Since the term was coined, the idea of the metaverse has remained more of a fictional concept than a scientific one. However, with technological advancements in recent years, the metaverse has become more tangible. Much of the recent hype happened after Mark Zuckerberg made the announcement to rename the Facebook brand to Meta. Many retailers have since jumped aboard the metaverse train.

Nike recently filed multiple trademarks allowing them to create and sell Nike shoes and apparel virtually. JP Morgan opened their first virtual bank branch. Samsung recreated their New York City flagship store in the virtual browser-based platform Decentraland, where they are launching new and creating events.

While many retailers are capitalizing on the metaverse early, there is still uncertainty about whether the metaverse really is the future of retailing or whether it will be a short-lived fad. ... (more that discusses this usefully)  ... 

Are Ads Causing Shoppers to Lose Faith in Amazon search results?

 Significantly?

Are ads causing shoppers to lose faith in Amazon search results?   by Matthew Stern in RetailWire.

Amazon.com has long been the main go-to place for online product search, but a recent Washington Post article finds that it is no longer giving customers what they want because advertisements are muscling out the real search results.

Running a search for “cat beds,” the Post found that the entire first screen of results displayed advertisements masquerading as listings rather than products arrived at because they provided the best combination of price and quality. The results included one ad that featured a dog in the picture, rather than a cat.

On the first five pages of search results, more than half of the listings were either ads or Amazon’s own products. A Profitero study found that Amazon lists sponsored products on its first page of search results at a rate twice that of Walmart and four times that of Target. Other online outlets, like food delivery platforms and Google and Apple’s app stores, have also started displaying ads as search results.

Amazon’s ad business grew 58 percent in 2021, making it the third largest online ad seller, according to a Recode article. Six high-volume sellers told Recode that to succeed on Amazon today, sellers have to spend between 10 and 20 percent of their sales on ads.

Amazon’s practices surrounding how it lists search results have been criticized before. ... ' 

From Handrwriting to Computer Code

Like to see a useful example of this, though who uses much handwriting anymore?    Do we need another introduction of uncertainty?

Programming Tool Turns Handwriting into Computer Code  By Cornell Chronicle, November 30, 2022

Writing code long-hand. 

The Notate pen-based interface lets users of computational, digital notebooks open drawing canvases and handwrite diagrams within lines of traditional, digitized computer code.

A team of Cornell University researchers created the Notate interface to translate handwriting and sketches into computer code.

The pen-based interface enables digital notebook users to open drawing canvases and to handwrite diagrams within lines of traditional code.

Notate is driven by a deep learning model, allowing notation in the handwritten diagram to reference textual code and vice versa.

Cornell's Ian Arawjo said, "People are ready for this type of feature, but developers of interfaces for typing code need to take note of this and support images and graphical interfaces inside code."

From Cornell Chronicle

View Full Article  

Existential Physics Reading

 Understanding physics and its practical application is becoming important.   Here a good intro I am reading,  notably about how quantum effects are being applied in computing and cryptography. 

Existential Physics: A Scientist's Guide to Life's Biggest Questions Audible Logo Audible Audiobook – Unabridged

By Sabine Hossenfelder (Author), Gina Daniels (Narrator), Penguin Audio (Publisher)

4.4 out of 5 stars    289 ratings

A contrarian scientist wrestles with the big questions that modern physics raises, and what physics says about the human condition.

Not only can we not currently explain the origin of the universe, it is questionable we will ever be able to explain it. The notion that there are universes within particles, or that particles are conscious, is ascientific, as is the hypothesis that our universe is a computer simulation. On the other hand, the idea that the universe itself is conscious is difficult to rule out entirely. 

According to Sabine Hossenfelder, it is not a coincidence that quantum entanglement and vacuum energy have become the go-to explanations of alternative healers, or that people believe their deceased grandmother is still alive because of quantum mechanics. Science and religion have the same roots, and they still tackle some of the same questions: Where do we come from? Where do we go to? How much can we know? The area of science that is closest to answering these questions is physics. Over the last century, physicists have learned a lot about which spiritual ideas are still compatible with the laws of nature. Not always, though, have they stayed on the scientific side of the debate. 

In this lively, thought-provoking book, Hossenfelder takes on the biggest questions in physics: Does the past still exist? Do particles think? Was the universe made for us? Has physics ruled out free will? Will we ever have a theory of everything? She lays out how far physicists are on the way to answering these questions, where the current limits are, and what questions might well remain unanswerable forever. Her book offers a no-nonsense yet entertaining take on some of the toughest riddles in existence, and will give the listener a solid grasp on what we know—and what we don’t know.

* This audiobook includes a downloadable PDF with key visual figures included in the book. 

PLEASE NOTE: When you purchase this title, the accompanying PDF will be available in your Audible Library along with the audio.

©2022 Sabine Hossenfelder (P)2022 Penguin Audio

Recent Game playing example

 

Games can be useful to test learning and alternative process options.

Why Researchers Are Teaching AI to Play Minecraft    By Popular Science, December 2, 2022

Nuclear power plant model made in Minecraft An artist's approximation of a nuclear power plant model made in Minecraft.  Using Video Pre-Training (VPT),the new AI program could construct items in Minecraft previously unattainable to bots reliant only on reinforcement learning.

OpenAI has developed a Minecraft-playing bot that can build pixelated tools and buildings in the game that require more than 20,000 consecutive actions via a combination of imitation and reinforcement learning. The bot, trained on 70,000 hours of human gameplay, is the first to build "diamond tools," which take human players 20 minutes and 24,000 actions, on average, to construct.

Imitation learning requires each step to be hand-labeled, but the researchers used a separate neural network to handle labeling via Video Pre-Training....  The researchers said the use of imitation and reinforcement learning in combination could pave the way for advancements in self-driving vehicles and nuclear fusion research.

From Popular Science    

View Full Article

Whats a Software BOM?

Billing it up for a Product

ACM NEWS

SBOM: An Up-Close Look at a Software Bill of Materials, November 30, 2022

Multiple Software Bill of Materials are produced at different levels of the software stack. 

A software bill of materials is a list of the base elements (such as code libraries) used to create a product.

Unless you've been living under a rock the past few years, you've likely at least heard of Log4j. This is an Apache open source library that's commonly used in just about everything Java-related online. Unfortunately, in late 2021 the logging package was discovered to be critically vulnerable to remote code execution attacks, meaning an attacker could exploit it to install malware (e.g., ransomware) onto vulnerable systems and inject larger networks.

Cloudflare CEO Matthew Prince reported on Twitter that there were 400 confirmed exploit attempts per second. But that's just one estimate — according to The Washington Journal, Akamai Technologies said it observed 10 million such exploit attempts per hour. Research from Check Point also showed that the attackers were rolling out new variants of the exploits — more than 0 in under 24 hours.

That's a lot of exploits and a lot of variations to boot. Considering that the Log4j vulnerability affected major companies like Amazon, Apple, and IBM, it's no surprise that it had companies globally worried.

But what makes the situation particularly concerning is that many companies weren't aware that the products they use contained such vulnerable elements. If only there was a way that organizations could know exactly what components are part of the software they use… Oh, wait, there is: they could use products that come with a software bill of materials (SBOM).

From hashedout by The SSL Store

View Full Article   

Coinbase and Apple

Fees for crypto transfers 

Coinbase removes NFT transfers from iOS app following dispute with Apple

By Duncan Riley in SiliconAngle

Coinbase Inc. is the latest company to have issues with Apple and its App Store policies, claiming that it has been forced to remove nonfungible token transfers from its iOS wallet because Apple wants a cut of the transfer fees.

Nonfungible tokens, better known as NFTs, rose to popularity in late 2020. Although interest in them has dropped precipitously, there is still an active market for buying and selling them. Most NFTs reside on the Ethereum blockchain and transfers on the blockchain require “gas,” a transaction fee, to make a transfer. It’s a fee that all users may pay to perform any function on the Ethereum blockchain.

Apple, for unclear reasons, now thinks it’s entitled to a 30% cut of the gas fee of any NFTs transferred using Coinbase. The case is unlike other disputes with Apple wherein developers, most notably Epic Games Ltd., argue that Apple’s 30% cut is unfair. At least when buying items in “Fortnite,” Apple is taking a cut out of Epic Games’ profits. There’s nothing to take from Coinbase because it’s not a fee it receives, keeps or profits from.

“For anyone who understands how NFTs and blockchains work, this is clearly not possible,” Coinbase wrote in a tweet. “Apple’s proprietary In-App Purchase system does not support crypto, so we couldn’t comply even if we tried.”  .... ' 

Laundry Folding Record

Addressing common tasks.

Robot Laundry Folding Speed Record Broken

Ars Technica, Benj Edwards, October 19, 2022

Researchers at the University of California, Berkeley and Germany's Karlsruhe Institute of Technology have achieved a new robot laundry-folding speed record with the SpeedFolding system. SpeedFolding combines machine vision, the BiManual Manipulation Network (BiMaMa-Net), and two industrial robot arms to fold 30 to 40 randomly positioned garments per hour; the researchers said the previous record was three to six folds per hour. The BiMaMa-Net neural network analyzed 4,300 human and machine-assisted examples to learn to fold clothing. The system uses an overhead camera to study the garment's initial state, and to calculate where to grasp it. The SpeedFolding robot can fold clothes from a random initial position in less than two minutes on average at a 93% success rate, as well as generalize to clothes of differing material, shape, or color than those it trained with.  .. ' 

Battery Supply Chain Comes to the US

Supply Chain elements.

The EV Battery Supply Chain Is Coming to America

Close-up of a mechanic's hands disassembling an electric car battery on top of a trailer inside a mechanic shop

November 16, 2022  Bloomberg

It’s official: the supply chain for electric vehicle batteries is coming to America.  

Redwood Materials Inc., the battery recycling company created by Tesla co-founder J.B. Straubel, said it has reached a deal to supply Panasonic with billions of dollars in critical battery components that will be produced in the U.S. for the first time.  

The agreement marks the first major contract for domestically processed cathode material, a substance that’s responsible for more than one third of the expense of a finished battery pack. The material will supply Panasonic’s new battery plant in Kansas City, Kansas, when mass production begins there in 2025. The plant is expected to produce cells primarily for Tesla electric vehicles.

“Panasonic has been a partner for many years, but this is very significant,” Straubel said in an interview. “This is a large portion of their cathode supply. It’s such an impactful announcement for the U.S. supply chain in general — and of course for us as a company.”

Straubel declined to specify the amounts of material to be supplied or the specific price structure, other than to say that it would total billions of dollars over several years.

A New U.S. Industry

Straubel left Tesla in 2019 after he grew concerned about a widening gap between electric vehicle demand and the availability of materials needed to make them. Redwood quickly rose to become the biggest lithium-ion battery recycler in the U.S., before branching out into anode and cathode production.

Every battery has two electrodes — a cathode and an anode — between which trillions of charged lithium atoms travel. It’s the cathode that largely determines a battery’s performance, cost and environmental footprint. Cathode today is produced almost entirely in Asia.

Straubel says the materials in Redwood’s cathode will help electric vehicles qualify for new $7,500 federal tax incentives available under stringent guidelines that will take effect between now and 2024.

To qualify for the full incentive under President Joe Biden’s 2022 Inflation Reduction Act, half of a battery’s minerals (by cost) must either be recycled in North America or mined from a country with which the U.S. has a free trade agreement. The cathode Redwood plans to provide Panasonic will be made with 100% recycled cobalt and 25% to 30% recycled nickel and lithium, Straubel said. He declined to say where the company would procure the remaining mineral supplies.

“We were already going at supersonic speeds,” Straubel said, but the manufacturing provisions in Biden’s climate plan were “like what happens if you attach a rocket engine to supersonic airplane.”

Cobalt is one of the most expensive and controversial ingredients for batteries. The majority of the world’s supply comes from the Democratic Republic of Congo, where allegations of human rights abuses in the mining industry have been common. The reason Redwood has such a large amount to recycle is that it’s used in much higher concentrations for batteries that power consumer electronics than for those used in EVs. For example, it would take 6,147 recycled iPhone batteries to provide enough lithium for a Tesla Model Y, but only 166 iPhones to provide enough cobalt, according to BloombergNEF data.

For now, most battery recycling consists of consumer electronics and scrap material from factories. That’s expected to change dramatically as the first fleets of mass-produced EVs continues to age.  

Panasonic’s new Kansas City plant will be the company’s second major battery factory in the U.S., after the Nevada Gigafactory it jointly operates with Tesla. Straubel said the $4 billion Kansas City plant is likely to eventually exceed Nevada production. Redwood already recycles the scrap materials produced at Panasonic’s Nevada factory, and in turn will supply that factory with refined materials, including anode copper foil, by the end of this year.

Enough for One Million Cars  .... ' 

European Automotive Does Quantum

Unexpected this early, but very interesting. 

The European automobile industry is going quantum in TheNextWeb  By Tristran Greene 

Spooky action at a distance goes vroom ... 

It’s a bold new world for automobile makers. After a century of development and fine-tuning, the combustion engine is going the way of the dodos as Europe shifts to clean energy.

But there’s more to the future of cars than just electric motors. The onset of fully autonomous vehicles may lie just beyond the technological horizon and the promise of a million-mile battery draws ever closer. In order to navigate the road to these technologies, European automakers are partnering with quantum computing companies at an increasing pace.

The European automobile industry has a long, rich history of technological innovation. From its onset with the Nesselsdorfer Wagenbau in 1898 to the masterpiece that is the 2023 McLaren Artura, Europe’s place at the cutting edge of the industry has never been questioned. With that in mind, let’s contemplate the future.

The next steps for the industry involve taking a quantum leap forward. Despite the fact that quantum computing and other quantum-based technologies are still in their infancy, there are myriad ways in which they can aid the automotive industry.

Right up front, the low hanging fruit is autonomous driving. Despite the early hype, researchers and automobile makers have yet to crack the self-driving car nut. For every step companies such as BMW, Tesla, and Waymo take forward, it seems like hundreds of edge cases pop up that the AI is unable to deal with.

We’re probably still a long way off from building a quantum computer that can fit into a car where, presumably, it would act as its brain. But quantum speedup — the ability for quantum processors to perform calculations and/or run algorithms that a classical system couldn’t do in a useful amount of time — could offer advances in several foundational areas for autonomous vehicle systems.

Scientists at Terra Quantum AG, recently partnered with Volkswagen to find novel methods for using hybrid quantum neural networks to improve image recognition. This particular experiment demonstrated the potential for quantum technologies to improve the quality assurance process drastically.

Essentially, the researchers used quantum-powered AI to increase the accuracy of its image detection abilities in order to improve the quality of the car manufacturing process. The techniques they’re working to develop could easily spill over into other industries, but they could also be used to give self-driving cars better “eyes” by increasing the speed and accuracy at which neural networks can process images.

Pasqal, a Paris-based quantum startup also partnered up with BMW in another quantum-based endeavor. Together with the German-owned automobile maker, the company hopes to find new, lighter, more durable materials to build cars out of. The team hopes to eventually reach the point where the design process is fast, accurate, and involves zero-prototyping in order to ensure a clean energy approach to every facet of the car-making process.

BMW and Volkswagen are early adopters out in front of the impending quantum computing hardware explosion, but you can be sure that every other major automobile maker also has a plan to get in on the action — experts predict the quantum technologies market will hit nearly $500B by 2030. And the shift towards autonomous vehicles (and away from ownership) will require an entirely different view on supply and logistics, something the quantum industry is heavily invested in improving.  ... '

On Swarm Robotics

Especially interested in how we can link to tasks that need multiples of specialized solvers ... 

Swarm Robotics Moves Forward   By Samuel Greengard

Communications of the ACM, December 2022, Vol. 65 No. 12, Pages 12-14   10.1145/3565979

Intelligent patterns and behaviors appear in many living systems—and the collective behavior of groups such as birds or insects frequently display greater intelligence than any given creature. What one ant, bee, or bird cannot tackle alone, the group can, and often does, accomplish.

This shared intelligence has not gone unnoticed by robotics researchers. As they look for ways to build smarter mechanical objects that can creep, run, swim, and fly, they are turning to nature for inspiration and direction. Drawing on areas as diverse as biology, psychology, physics, materials science, and aerodynamics, they are exploring ways to create new and smarter robotic systems, including microbots that could someday venture into the human body.

"The field is very important, especially for large-scale applications such as environmental monitoring, search and rescue, agriculture, exploration, and crisis response," says Katia P. Sycara, the Edward Fredkin Research Professor of Robotics at Carnegie Mellon University. "A swarm is a large group of robotic devices—into the hundreds or thousands—that can perform various useful large-scale tasks. Swarms can also remain robust with the loss of individual members."

Of course, swarm robotics is not without technical and practical challenges—and it introduces new risks, such as the ability to use systems in wars and for autonomous killing. Yet, recent research demonstrates the concept is increasingly viable. "In many task settings, large solitary robots are about to give way to robot collectives capable of more complex behaviors," says Kirstin Petersen, an assistant professor of electrical and computer engineering in the College of Engineering of Cornell University.

Second Nature

Drawing inspiration from nature is at the center of swarm robotics. "If you examine an army ant, each one is very simple and can do very little on its own," observes Steven Ceron, a Ph.D. candidate and researcher in the mechanical engineering department at Cornell University. "However, when you look at what an entire colony can do, it's pretty remarkable."

"In many task settings, large solitary robots are about to give way to robot collectives capable of more complex behaviors."

For example, a group of ants can work collectively to build a bridge with their bodies to allow other ants to cross over a stream. Similarly, bees communicate with each other about food sources that allow a colony to flourish, and many types of migratory birds fly in formations that benefit the group. The birds at the front reduce the drag for others behind them and, periodically, the birds switch positions so no single bird bears the entire brunt of the workload. "The collective group has an intelligence that exceeds any individual animal," Ceron says.  ... ' 

3D Printing Can Manufacture Customized Sensors

 Printing Sensors

3D Printing Can Manufacture Customized Sensors

KTH Royal Institute of Technology (Sweden)

David Callahan, September 28, 2022

A new three-dimensional (3D) printing method could cost-effectively fabricate insect-sized electronic sensors for robots and medical devices. Scientists at Sweden's KTH Royal Institute of Technology led the technique's development to overcome the limits of conventional microelectromechanical system (MEMS) manufacturing. The researchers enhanced two-photon polymerization, a process that can generate high-resolution objects hundreds of nanometers in size, to sense functionality. The method forms transducing elements via stencil-like shadow-masking. KTH's Frank Niklaus said the approach can produce about a dozen customized MEMS accelerometers in a few hours using commercial manufacturing tools. "The new capabilities offered by 3D-printed MEMS could result in a new paradigm in MEMS and sensor manufacturing," said Niklaus. ... ' 

On Ray and Path Tracing

Representing 3D Objects.

What Is Path Tracing?

March 23, 2022 by Brian Caulfield

Turn on your TV. Fire up your favorite streaming service. Grab a Coke. A demo of the most important visual technology of our time is as close as your living room couch. Propelled by an explosion in computing power over the past decade and a half, path tracing has swept through visual media. It brings big effects to the biggest blockbusters, casts subtle light and shadow on the most immersive melodramas and has propelled the art of animation to new levels.

Path tracing is going real time, unleashing interactive, photorealistic 3D environments filled with dynamic light and shadow, reflections and refractions.

So what is path tracing? The big idea behind it is seductively simple, connecting innovators in the arts and sciences over the span half a millennium.

What’s the Difference Between Rasterization and Ray Tracing?

First, let’s  define some terms, and how they’re used today to create interactive graphics — graphics that can react in real time to input from a user, such as in video games.

The first, rasterization, is a technique that produces an image as seen from a single viewpoint. It’s been at the heart of GPUs from the start. Modern NVIDIA GPUs can generate over 100 billion rasterized pixels per second. That’s made rasterization ideal for real-time graphics, like gaming.

Ray tracing is a more powerful technique than rasterization. Rather than being constrained to finding out what is visible from a single point, it can determine what is visible from many different points, in many different directions. Starting with the NVIDIA Turing architecture, NVIDIA GPUs have provided specialized RTX hardware to accelerate this difficult computation. Today, a single GPU can trace billions of rays per second.

Being able to trace all of those rays makes it possible to simulate how light scatters in the real world much more accurately than is possible with rasterization. However, we still must answer the questions, how will we simulate light and how will we bring that simulation to the GPU?

What’s Ray Tracing? Just Follow the String

To better answer that question, it helps to understand how we got here.

David Luebke, NVIDIA vice president of graphics research, likes to begin the story in the 16th century with Albrecht Dürer — one of the most important figures of the Northern European Renaissance — who used string and weights to replicate a 3D image on a 2D surface.

Dürer made it his life’s work to bring classical and contemporary mathematics together with the arts, achieving breakthroughs in expressiveness and realism.

The string’s the thing: Albrecht Dürer was the first to describe what’s now known as “ray tracing,” a technique for creating accurate representations of 3D objects on a 2D surfaces in Underweysung der Messung (Nuremberg, 1525),f15

In 1525 with Treatise on Measurement, Dürer was the first to describe the idea of ray tracing. Seeing how Dürer described the idea is the easiest way to get your head around the concept.

Just think about how light illuminates the world we see around us. Now imagine tracing those rays of light backward from the eye with a piece of string like the one Dürer used, to the objects that light interacts with. That’s ray tracing.

Ray Tracing for Computer Graphics    .... 

Opinion on the CHIPS Act

 Short Opinion piece from ACM site.  Have not reviewed further, but plan to

The False Promise of America's CHIPS Act

By Project Syndicate, November 28, 2022

The CHIPS Act is flawed in many ways.  (Opinion) 

The U.S. CHIPS and Science Act allocates more than $50 billion to strengthen the semiconductor industry in the hope of making the U.S. self-sufficient. But it is doubtful that the CHIPS Act will achieve its stated goals, much less that it should be used as a model for similar support to other industries.

While it is understandable that the U.S. wants to maintain its technological lead in chips, subsidizing large existing firms and government management of the industry almost certainly will not achieve that goal.

From Project Syndicate

View Full Article  

AWS Wants Cloud in Space

 New domains, one answer to supply chains there. 

AWS sets its sights on space for the future of cloud capabilities  By Mike Moore published about 3 hours ago

AWS wants to make the world a better place, from space

Cloud computing has revolutionized all kinds of business and workplace environments, but one of the biggest players in the industry now hopes it can help accelerate the next big breakthroughs in space as well.

Speaking at the recent AWS re:Invent 2022 event, Clint Crosier, the company’s director of aerospace and satellite, outlined how the company is heavily backing the use of cloud computing in space.

At a panel session alongside Peggy Whitson, astronaut and director of human space flight at Axiom, Crosier described, “what we at AWS call the making the world a better place from space mission.”  ... ' 

On Event Modeling

 Its all about events and how we handle them. 

Adam Dymitruk on Event Modeling    By Software Engineering Radio    November 28, 2022

Credit: Adam Dymitruk/Software Engineering Radio

Adam Dymitruk is CEO and founder of Adaptech Group.

In an interview, Dymitruk explores the event-modeling approach to discovering requirements and designing software systems. Dymitruk explains how the structured approach eliminates the specifics of implementation details and technology decisions, enabling clearer communication for all stakeholders while keeping conversations focused on the business opportunity.

Using concrete examples of event modeling in practice, the interview examines event modeling in the context of other related approaches and methodologies, including event sourcing, event storming, CQRS, and domain-driven design.

From Software Engineering Radio

View Full Article     

Data Intelligence Platform Seek AI Launches to Automate Repetitive Tasks

 Automate and make intelligent. 

Data Intelligence Platform Seek AI Launches to Automate Repetitive Tasks

In DataNami  by Jaime Hampton

With all the hype around generative AI and its text-to-image capabilities throwing the art world for a loop, it may be easy to forget that you can write code with it, too.

Seek AI, a new data intelligence platform that uses generative AI, specifically large language models, has just been launched. The company is the creation of CEO Sarah Nagy, a data professional and entrepreneur who felt the available data tools were not adequate for helping her less technical colleagues access needed data. Nagy and co-founder Sarah Smith decided to look to these newest advances in AI for a solution.

“For anyone that has worked with generative AI like DALL-E, Stable Diffusion, or GPT-3, you have experienced the power of these models, which work like magic. But did you know that generative AI can also write code? When I first saw this in action, I knew instantly that it would change the way people would work with data,” Nagy said in a company mission statement.

Data access can be challenging for business users less familiar with coding, which can add hours of work to the already heavy workflows of the data professionals that assist them. As part of her data career, Nagy says she spent a significant amount of time each day writing manual code by hand to pull data for her colleagues.

Seek AI has launched its SaaS platform to automate these repetitive tasks. According to the company, Seek AI allows data teams to automate and oversee database query projects to improve productivity, with a focus on sales and marketing purposes.

“We’ve built a platform that uses artificial intelligence to reduce inefficiencies in accessing data,” says Nagy. “Currently, data scientists and business analysts waste time manually writing repetitive code. We’ve automated that process with sophisticated natural language processing and machine learning.”

The complex deep-learning models behind OpenAI’s DALL-E and GPT-3 are also behind the Seek AI platform’s technology and form the basis for its natural language interface that allows users to ask questions and obtain immediate answers. The models understand natural language commands and generate the necessary code for instantly querying databases.

Seek AI’s NLP interface transforms natural language questions into code for querying databases. Source: Seek AI

Seek AI’s natural language interface can be accessed through email, Slack, text, and several CRM systems. Available as a monthly or per user subscription, the company says its platform allows data teams to automate the generation and maintenance of code for answering ad-hoc questions, as well as generating and maintaining semantic models.

“By helping to remove some of the mundane coding and waiting from both the data team and the business-facing teams, I hope that Seek will make knowledge workers happier because they will be able to pursue more meaningful work. Not to mention the value it adds to organizations by allowing them to make better use of the data they work so hard to compile and maintain,” said Nagy.

“Giving fast, high-quality answers to our customers’ questions is the difference between winning and losing over our competitors,” said Tim Harrington, CEO of financial dataset supplier Battlefin, in a release. “Seek AI played a critical role in our company’s 2023 strategy because of the edge that it gives us in accessing and analyzing our 2,400+ datasets in response to customer questions. I’d estimate that our ROI on Seek AI is about 10x based on what we would have spent to achieve this level of efficiency without the platform.”  .... ' 

Mechanical Neural Networks

Quite technical, but worth a link for later study.... 

Mechanical neural networks: Architected materials that learn behaviors  ... 

BY RYAN H. LEE ERWIN A. B. MULDER ET AL.

One for all: Universal material model based on minimal state-space neural networks

Abstract

Aside from some living tissues, few materials can autonomously learn to exhibit desired behaviors as a consequence of prolonged exposure to unanticipated ambient loading scenarios. Still fewer materials can continue to exhibit previously learned behaviors in the midst of changing conditions (e.g., rising levels of internal damage, varying fixturing scenarios, and fluctuating external loads) while also acquiring new behaviors best suited for the situation at hand. Here, we describe a class of architected materials, called mechanical neural networks (MNNs), that achieve such learning capabilities by tuning the stiffness of their constituent beams similar to how artificial neural networks (ANNs) tune their weights. An example lattice was fabricated to demonstrate its ability to learn multiple mechanical behaviors simultaneously, and a study was conducted to determine the effect of lattice size, packing configuration, algorithm type, behavior number, and linear-versus-nonlinear stiffness tunability on MNN learning as proposed. Thus, this work lays the foundation for artificial-intelligent (AI) materials that can learn behaviors and properties. ... ' 

INTRODUCTION

Scientists have been inspired by the interconnected network of neurons that constitute biological brains and enable complex learning with unmatched speed and energy efficiency. Consequently, many have sought to leverage a variety of interconnected networks to mimic natural learning for numerous artificial-intelligent (AI) applications (1–3).

Some of the first networks developed for AI purposes were purely mathematical in form. The concepts underlying these mathematical networks, called artificial neural networks (ANNs) (Fig. 1A), were first introduced by McCulloch and Pitts (4) but were later matured by Rosenblatt (5). The mathematical formulation underlying ANNs can be diagrammed using interconnected lines, shown in blue in Fig. 1A, that represent scalar values, called weights (6), which are multiplied by input numbers that are fed into multiple layers of activation functions (6), called neurons, which ultimately produce output values. If the ANN is provided with a set of known input and output values, then the network can be trained by tuning its weights so that it accurately predicts previously unknown output values that result for any desired input values. Hornik et al. (7) proved the true AI potential of ANNs by demonstrating that, with sufficiently large numbers of neurons and layers, ANNs could learn to model almost anything by accurately mapping any number of inputs to any number of outputs. Tuning the weights of sizeable ANNs, however, proved to consume large amounts of computational time and energy using traditional digital computers.  ... ' 

Synthetic Data Can Offer ML Performance Improvements

 Previously mentioned, considering use of idea in test.

MIT News, Adam Zewe, November 3, 2022

A team of researchers at the Massachusetts Institute of Technology (MIT), the MIT-IBM Watson AI Laboratory, and Boston University found synthetic data can improve machine learning (ML) model performance. The researchers amassed the Synthetic Action Pre-training and Transfer dataset of 150,000 video clips to train three ML models on a wide range of human actions. They found the three models outperformed models trained with real video clips on four of six datasets, yielding the highest accuracy for datasets featuring clips with "low scene-object bias." MIT-IBM Watson AI Lab's Rogerio Feris said, "The ultimate goal of our research is to replace real data pretraining with synthetic data pretraining." Feris said that while there is a cost in creating an action in synthetic data, “Once that is done, then you can generate an unlimited number of images or videos by changing the pose, the lighting, etc."  ... ' 

Choosing Best Machine Learning Algorithm

 Simplified example

DataScienceDojo (@Data Science Dojo) Tweeted: 🤔 Here is how to choose the correct machine-learning algorithm!

Source: datasciencebrain

https://twitter.com/DataScienceDojo

#MachineLearning #DataScience https://twitter.com/datasciencedojo/status/1588715224605106176?s=51&t=e1T9N4c2nhYKMsgxSF1GFQ

Disney Neural Network Can Change an Actor's Age with Ease

Powerful idea,  So do I get to own all my transformed RE-ages?  Lighting conditions?  I am seeing some legal implications. 

Disney Neural Network Can Change an Actor's Age with Ease

Ars Technica, Benj Edwards, November 30, 2022

A neural network developed by researchers at Disney offers a cheaper, faster way to make actors look older or younger in TV or film. Disney said the artificial intelligence technique called FRAN (Face Re-aging Network) is "the first practical, fully automatic, and production-ready method for re-aging faces in video images." The researchers trained FRAN to learn how a person's appearance changes with age by feeding it thousands of images of faces synthetically aged from 18 to 85 years old by StyleGAN2. Disney said the “production-ready” system is the first able to age an actor’s image in various expressions, lighting conditions, and viewpoints. ... '

We Will See a Completely New Type of Computer, Says AI Pioneer Hinton

 Hinton predicts: A 'mortal' Neuromorphic Computer?  Ready to sign up and test.   When? 

We Will See a Completely New Type of Computer, Says AI Pioneer Hinton

In ZDNet, Tiernan Ray, December 1, 2022

Artificial intelligence pioneer and 2018 ACM A.M. Turing award recipient Geoffrey Hinton envisions a "mortal" neuromorphic computer combining hardware and software. Speaking at the Neural Information Processing Systems conference, Hinton said mortal computation means "the knowledge that the system has learned and the hardware, are inseparable." Hinton said such computers could be grown, forgoing costly chip fabrication, and he imagines they will be "used for putting something like GPT-3 in your toaster for $1, so running on a few watts, you can have a conversation with your toaster." He suggested a forward-forward neural network model, eliminating the backpropagation common to most neural networks, might suit mortal computation hardware.

Article

Race to the Moon?

 I continue to be interested in the supply chain processes involved.

China and NASA are racing to the moon. Side-by-side photos hint NASA has the edge, but China's secrecy makes the race hard to call.    By Morgan McFall-Johnsen Nov 30, 2022 in BusinessInsider

two rocket launches at night split image with Long March-2F on the left and nasa space launch system on the right

China's Shenzhou 15 mission, left, lifted off on Tuesday, one week after NASA launched its Artemis I mission to the moon, right. CNS/AFP/Getty Images; NASA/Joel Kowsky

China and NASA are racing toward the first human moon landing in 50 years.

Two recent launches suggest NASA has an edge, but China could be close behind.

Side-by-side images show how NASA and China are overlapping in the new space race for the moon. ... ' 

Fusion Power Still Years Away

Years Away most followers agree ... 

Fusion power is 'approaching' reality thanks to a magnetic field breakthrough

A boost from magnetism is nearly enough to achieve fusion ignition.

By Jon Fingas|@jonfingas|November 30, 2022 2:52 PM

Fusion power may be a more realistic prospect than you think. As Motherboard reports, researchers at the Energy Department's Lawrence Livermore National Laboratory have discovered that a new magnetic field setup more than tripled the energy output of the fusion reaction hotspot in experiments, "approaching" the level required for self-sustaining ignition in plasmas. The field was particularly effective at trapping heat within the hotspot, boosting the energy yield.

The hotspot's creation involved blasting 200 lasers at a fusion fuel pellet made from hydrogen isotopes like deuterium and tritium. The resulting X-rays made the pellet implode and thus produce the extremely high pressures and heat needed for fusion. The team achieved their feat by wrapping a coil around a pellet made using special metals.

The notion of using magnets to heat the fuel isn't new. University of Rochester scientists found they could use magnetism to their advantage in 2012. The Lawrence Livermore study was far more effective, however, producing 40 percent heat and more than three times the energy.

Practical fusion reactors are still many years away. The output is still far less than the energy required to create self-sustaining reactions. The finding makes ignition considerably more achievable, though, and that in turn improves the chances of an energy-positive fusion system. This also isn't the end of the magnetism experiments. A future test will use an ice-laden cryogenic capsule to help understand fusion physics. Even if ignition is still distant, the learnings from this study could provide a clearer path to that breakthrough moment.  ... ' 

Breaking the Scaling Limits of Analog Optical Computing

 Reducing error in optical networks 

Breaking the Scaling Limits of Analog Computing   By MIT News, November 30, 2022

Using the new technique, the larger an optical neural network becomes, the lower the error in its computations. 

Massachusetts Institute of Technology researchers have developed a technique that greatly reduces the error in an optical neural network.

As machine-learning models become larger and more complex, they require faster and more energy-efficient hardware to perform computations. Conventional digital computers are struggling to keep up.

An analog optical neural network could perform the same tasks as a digital one, such as image classification or speech recognition, but because computations are performed using light instead of electrical signals, optical neural networks can run many times faster while consuming less energy.

However, these analog devices are prone to hardware errors that can make computations less precise. Microscopic imperfections in hardware components are one cause of these errors. In an optical neural network that has many connected components, errors can quickly accumulate.

Even with error-correction techniques, due to fundamental properties of the devices that make up an optical neural network, some amount of error is unavoidable. A network that is large enough to be implemented in the real world would be far too imprecise to be effective.

From MIT News   View Full Article     

An AI-first Approach to Accelerating Autonomous Trucking

Of interest, December 7 

Reminder: December 7 Talk, "An AI-first Approach to Accelerating Autonomous Trucking" with Raquel Urtasun, Founder and CEO of Waabi

If you haven't done so yet, register now  for the next free ACM TechTalk, "An AI-first Approach to Accelerating Autonomous Trucking," presented on Wednesday, December 7 at 12:00 PM ET/17:00 UTC by Raquel Urtasun, Founder and CEO of Waabi, Professor at the University of Toronto, and a co-founder of the Vector Institute for AI. The co-editors of the ACM Journal on Autonomous Transportation Systems, Satish Ukkusuri and Vaneet Aggarwal, of Purdue University, will moderate the questions and answers session following the talk.

Leave your comments and questions with our speaker now and any time before the live event on ACM's Discourse Page.     And check out the page after the webcast for extended discussion with your peers in the computing community, as well as further resources on AI and autonomous vehicles.

(If you'd like to attend but can't make it to the virtual event, you still need to register to receive a recording of the TechTalk when it becomes available.)

Note: You can stream this and all ACM TechTalks on your mobile device, including smartphones and tablets.

Founded in 2021 by world-leading AI pioneer, Raquel Urtasun, Waabi is a next-generation autonomous vehicle company with a radically new AI-first approach to autonomous driving, starting with the trucking industry. In this talk, Raquel will dive into the ground-breaking AI-first approach needed to accelerate autonomy, and what it takes to bring the promise of self-driving closer to commercialization than ever before.

Duration: 60 minutes (including audience Q&A)

Presenter:   Raquel Urtasun, Founder and CEO, Waabi

Raquel Urtasun is the Founder and CEO of Waabi, a Full Professor in the Department of Computer Science at the University of Toronto, and a co-founder of the Vector Institute for AI. She is a world leading expert in AI for self-driving cars. Her research interests include ML, computer vision, robotics, AI, and remote sensing. Her lab was selected as an NVIDIA NVAIL lab. Urtasun is a recipient of an NSERC EWR Steacie Award, an NVIDIA Pioneers of AI Award, a Ministry of Education and Innovation Early Researcher Award, three Google Faculty Research Awards, an Amazon Faculty Research Award, two NVIDIA Pioneer Research Awards, a Connaught New Researcher Award, a Fallona Family Research Award, and two Best Paper Runner up Prize awarded at CVPR in 2013 and 2017 respectively. She was also named Chatelaine 2018 Woman of the year and one of Toronto's 2018 top influencers by Adweek magazine. She was a panelist at ACM's 75th Anniversary Celebration in 2022.

Moderator: Satish Ukkusuri, Professor, Purdue University; Co-editor of the ACM Journal of Autonomous Transportation Systems

Satish Ukkusuri is a Reilly Professor in the Lyles School of Civil Engineering at Purdue University and Director of the Urban Mobility Networks and Intelligence (UMNI) Lab. He is a University Faculty Scholar (2017-present), ASCE Fellow, Fulbright Fellow, a selectee of the National Academy of Engineering (NAE) JAFOE conference (2016), a selectee of the National Academy of Science (NAS) Arab American Frontiers of Science, Engineering and Medicine in 2017 and a CUTC/ARTBA Faculty Award (2011) among other awards. He has published more than 350 peer reviewed papers in journals such as PNAS, Nature Communications, and Transportation Research Part B. He is the Editor in Chief of the Journal of Big Data Analytics in Transportation and co-Editor in Chief of the ACM Journal of Autonomous Transportation.

Mexico Replacing China for Manufacturing?

Mentioned in passing in a meeting, this related Podcast of interest:

THE SCB PODCAST

Podcast | Can Mexico Replace China as a Manufacturing Powerhouse?

China is losing its luster as a low-cost source of manufacturing. Is Mexico an alternative?

U.S. manufacturers are beginning to sour on China as the place to go for cheap production. Reasons include the rising cost of Chinese labor, the distance to American markets and simmering trade tensions between the two countries. Can Mexico step up as companies think about diversifying sourcing and near-shoring production? On this episode, we hear the case for Mexico as the place to site manufacturing from Jay Gerard, head of customs with Nuvocargo. He discusses the new opportunities that exist in cross-border trade, but he’s also candid about the challenges that must be overcome if Mexico is to solidify its position as the ideal location for the production of U.S.-bound consumer goods. Are companies listening?

Hosted by Bob Bowman, Editor-in-Chief of SupplyChainBrain.

Show notes:  A blog post from Nuvocargo: 

Would You Like a QR Code in That Cookie? Food and 3D Printing

Waste Reduction embedded, 3D printed.

Would You Like a QR Code in That Cookie?  By AlphaGalileo, October 19, 2022

Given that food tags and packaging are a large source of waste worldwide, this new method of embedding edible information into food also will be important for waste reduction.

A three-dimensional (3D) printing method developed by scientists at Japan's Osaka University can embed edible quick response (QR) codes inside food.

"Many foods can now be produced using 3D printers," explained Osaka University's Yamato Miyatake.

"We realized that the insides of edible objects such as cookies could be printed to contain patterns of empty spaces so that, when you shine a light from behind the cookie, a QR code becomes visible and can be read using a cellphone."

The "interiqr" technique produces a QR tag made from the cookie itself, keeping taste, flavor, and outward appearance intact.

From AlphaGalileo

View Full Article   

How Will the Space Economy Change the World?

 Good overview of Space and Economy. Not enough regards the timing and risks involved.  And how will key aspects, like supply chains,  be effectively translated into Space?  Note we need to do a key materials analysis to determine needs and sourcing,   Contact me for thoughts on Space Supply Chains. 

How will the space economy change the world?

November 28, 2022 | Article By Ryan Brukardt   in McKinsey

Space is no longer the sole domain of governments and aerospace and defense companies. Businesses that pursue emerging opportunities now may gain a first-mover advantage.    Article (6 pages) 

The passengers who boarded commercial flights just after World War II didn’t know that air travel would begin to soar over the next decade, nor did the masses who first logged onto the internet in the 1990s realize that computers would one day provide much of their news, entertainment, and social life. And today, few people understand that the space economy—broadly defined as activities in orbit or on other planets that benefit human beings—could soon transform how they live and work.

Some hints of the coming changes are apparent, including the frequent headlines about SpaceX, Blue Origin, and other private companies launching their own rockets and deploying satellite constellations. These activities, once primarily the domain of government agencies, are now possible in the private sector because recent technological advances in manufacturing, propulsion, and launch have made it much easier and less expensive to venture into space and conduct missions. Lower costs have opened the door to new start-ups and encouraged established aerospace companies to explore novel opportunities that once seemed too expensive or difficult. The technological improvements have also intrigued investors, resulting in a surge of space funding over the past five years.

The potential for innovative space applications is immense, especially if established aerospace companies form partnerships with businesses that traditionally haven’t ventured into orbit. Pharmaceutical companies might establish a lab on a space station to study cell growth, for instance, or semiconductor companies might manufacture chips in extraterrestrial factories to determine whether any aspects of the space environment, such as the lack of gravity, improve the process. Such possibilities, which might have seemed like the stuff of science fiction a few years ago, could become an essential part of a business across multiple industries in the near future.

But how and when should companies take advantage of their greater access to space and pursue emerging use cases? And how can they decide what opportunities are most promising when the technology is so nascent? Although much remains uncertain, companies that begin exploring these questions now could gain a long-term advantage.

The benefits of the space economy—with more to come

Space has long been a potent incubator for innovation—first from governments and large telcos and now from multiple private companies as well. From the launch of Sputnik 1 in 1957 through today, the space economy has delivered most of its value through satellite services, including communications and data and image collection and analysis. Satellites help large companies with multiple tasks, including inventory monitoring at distant locations, instant authorization of credit-card transactions, and international videoconferencing. Consumers use satellite technology whenever an online navigation system pinpoints their location, or when they make calls during plane flights or from rural locations that lack cell phone towers. And television viewers can thank satellites for beaming the signals that allow them to watch their favorite programs. The role of satellites in these activities is often overlooked—many people may think terrestrial computer networks provide the necessary connectivity—unless a glitch occurs and draws attention to the unobtrusive technology operating in the background.

In addition, satellites help world leaders address intractable social, environmental, and economic challenges. Consider a few ways that satellite data can provide insights: ... '

Quantum Computers and Liquid-Like Electrons

 New idea to me.

Quantum Computing May be Bolstered by Liquid-Like Electrons,  By Tom's Hardware

November 1, 2022

When electrons are made to act in what’s known as a “helical Tomonaga-Luttinger liquid,” there are fewer particle interactions and energy exchange between them and the system. Experiments conducted by scientists at the Nanyang Technological University in Singapore may strengthen quantum computing with liquid-like electrons called parafermions. Electrons coaxed to act in a helical Tomonaga-Luttinger liquid exchange fewer particles and energy between them and the system, reducing the likelihood of error and decoherence in quantum systems.

The researchers deposited atom-thick crystals of tungsten ditelluride onto a graphene substrate, creating a quantum spin Hall insulator.

They cooled the substrate toward absolute zero degrees Fahrenheit (-273 degrees Celsius) and observed increased electron repulsion and collective motion where parafermions theoretically exist.

The researchers intend to lower temperatures to 150 millikelvins for experiments that should enable actual observation of parafermion clusters.

From Tom's Hardware    View Full Article   

New Book on Synthetic Data

The data is most often,  the thing.    Look forward to taking a look.

New Book: Synthetic Data

Published on November 28, 2022, By Vincent Granville

Author of Intuitive Machine Learning and Explainable AI.

84 articles

Synthetic data is used more and more to augment real-life datasets, enriching them and allowing black-box systems to correctly classify observations or predict values that are well outside of training and validation sets. In addition, it helps understand decisions made by obscure systems such as deep neural networks, contributing to the development of explainable AI. It also helps with unbalanced data, for instance in fraud detection. Finally, since synthetic data is not directly linked to real people or transactions, it offers protection against data leakage. Synthetic data also contributes to eliminating algorithm biases and privacy issues, and more generally, to increased security.

Terrain generation, evolution and morphing (in chapter 11)

This book is the culmination of years of research on the topic, by the author. Not only it integrates all the material from his previous book “Intuitive Machine Learning and explainable AI”, but it also contains all but the most advanced math from his book on stochastic simulations. The author also added more recent advances with applications to terrain generation (with animated data), synthetic universes and experimental math. The latter is an infinite source of synthetic data to build and benchmark new machine learning techniques. Conversely mathematics benefits from these techniques to uncover new insights related to the most famous unsolved math problems. Chapter 14 on the Riemann Hypothesis illustrates this point, with new state-of-the-art research results on the subject.  ... ' 

Predictive Database for Materials

Materials science becoming increasingly important

 Nanoengineers Develop Predictive Database for Materials

UC San Diego Today, Emerson Dameron, November 28, 2022

The M3GNet algorithm developed by nanoengineers at the University of California, San Diego (UCSD)'s Jacobs School of Engineering can forecast the structure and dynamic properties of any material almost instantaneously. Researchers used M3GNet to compile the matterverse.ai database of more than than 31 million yet-to-be-synthesized materials with traits predicted by machine learning algorithms. UCSD's Shyue Ping Ong and colleagues combined graph neural networks with many-body interactions into a highly accurate deep learning framework that operates across the entire periodic table. The team employed the Materials Project's database of materials energies, forces, and stresses to train the predictive M3GNet interatomic potential model. "We truly believe that the M3GNet architecture is a transformative tool that can greatly expand our ability to explore new material chemistries and structures," said Ong.

Full Article    

San Francisco to allow police 'killer robots'

 Depending on the definition, doubt this. 

San Francisco to allow police 'killer robots'   in the BBC

By Ben Derico & James Clayton,  BBC News, San Francisco

San Francisco's ruling Board of Supervisors has voted to let the city's police use robots that can kill.

The measure permits police to deploy robots equipped with explosives in extreme circumstances. Dr Catherine Connolly, from the group Stop Killer Robots, said the move was a "slippery slope" that could distance humans from killing.The city's police - the SFPD - told the BBC they do not currently operate any robots equipped with lethal force. They said though that there may be future scenarios in which lethal force could be used by a robot.

A spokesperson for the police said "robots could potentially be equipped with explosive charges to breach fortified structures containing violent, armed, or dangerous subjects".They also said robots could be used to "incapacitate, or disorient violent, armed, or dangerous suspects who pose a risk of loss of life".

Advocates for the measure said it would only be used in extreme situations. Opponents, however, say the authority could lead to further militarisation of the police force.

The measure passed, with an amendment on Tuesday specifying that officers could only use robots wielding deadly force after employing alternative de-escalation tactics. The board also stipulated that only a limited number of high-ranking officers could authorise its use. This type of lethal robot is already in use in other parts of the United States.

In 2016, police in Dallas, Texas, used a robot armed with C-4 explosive to kill a sniper who had killed two officers and injured several more. The SFPD said the department does not currently own any robots outfitted with lethal force, but said the measure might be needed in the future.  "No policy can anticipate every conceivable situation or exceptional circumstance which officers may face. The SFPD must be prepared, and have the ability, to respond proportionally," a spokesperson said.

The federal government has long dispensed military grade equipment, camouflage uniforms, bayonets and armoured vehicles to help local law enforcement.  But a California state law passed this year now requires city police forces to inventory military-grade equipment and seek approval for their use. Dr Catherine Connolly, from the campaign group Stop Killer Robots, said the move could "make humans more and more distant from the use of force and the consequences of the use of force". She also said the measure could make it "easier to make decisions to use lethal force in the first place".

Alexa Announces an A/B Testing Service

Could have used this.

Announcing the general availability of the Alexa A/B testing service

Staff Writer Sep 30, 2022

The Alexa Skills A/B testing service, which allows skill builders to make informed design choices is generally available today

In the third quarter of 2021, the development team at Vocala had to make a critical decision: they had to determine the ideal length of a prompt that would motivate users to subscribe to the paid version of their skill.

Vocala is a leading voice design studio based in the United Kingdom. Over the last few years, Vocala has developed Interflora, the United Kingdom’s first retail sector voice-based skill, in addition to developing a recruiting skill for the British Royal Navy. Vocala is also the publisher of the Yak Yak Games Skill that has a 4.5 rating on the Amazon UK store. Yak Yak Games is a curated collection of leading games such as Deal or No Deal, Song Blast, and Jeff Stelling's Sports Quiz. The skill has a seven-day free trial, at the end of which customers can sign up for a paid version with a monthly paid subscription.

“We wanted to test the prompt that would result in the greatest number of paid conversions,” says James Holland, lead voice developer at Vocala. “More specifically, we wanted to run an A/B test to determine whether a longer or shorter prompt would help us achieve our goal.”

A/B testing is a commonly used experimentation process to make design-related decisions. For web or mobile design, these variables could include the placement of a form on the page, or the color of a Register button. For skill builders, A/B testing can be utilized to arrive at the optimal utterances in order to achieve objectives such as reduced customer friction, increased daily repeat visits, and greater revenue.

Historically, designing and deploying A/B tests for Alexa skills has been a cumbersome process. Setting up infrastructure for an A/B test could take as long as two weeks.

“Working on multiple projects at one time, and responding to clients who want quick turnarounds, we don’t always have the bandwidth for lengthy A/B experiments,” said Holland.

As part of their efforts to continually improve Alexa’s offerings for developers, Alexa solution architects meet regularly with skill builders like Vocala. In response to recent developer feedback related to the importance of experimentation, today, Amazon is launching the Alexa Skill A/B testing service. The new service allows skill builders to design A/B experiments with the goal of maximizing in-skill purchases, repeat visits, and number of dialogs for a session. ... ' 

Defense Industry Challenges

The business of defending,

This is the first in a series of articles about the inflation challenge within the defense industry.

Inflation continues to run above historical averages. The general public is feeling the pinch, with the consumer price index up 8.3 percent over the past year.1 Within the aerospace and defense sector, leaders are trying to adjust to an inflation level that many have never encountered in their careers.

The producer price index (PPI), which measures the prices received by domestic US businesses for their output, has increased 12.2 percent for goods and 6.8 percent for services over the past year.2 Input costs for common defense products and services have risen even higher, with increases of 10.0 percent for industrial chemicals, 17.8 percent for transportation and warehousing, and 25.9 percent for energy.

Rising costs may ultimately cut into the US Department of Defense’s (DOD’s) buying power. In an earlier analysis, we estimated that the agency could lose more than $100 billion in purchasing power within five years if the economy reenters a period of high inflation and the defense budget is limited to nominal top-line increases, similar to what happened in the 1970s.3 In that scenario, the DOD would have limited funds to invest in modernizing military equipment, especially since the department would still have to increase military pay to meet mandatory requirements. Defense companies would then face lower demand on top of increased input costs.

Today’s soaring prices come after many years of relatively low inflation and have prompted the US Federal Reserve to raise interest rates more rapidly than any time in the past two decades.4 While many businesses are also revisiting their strategies in response to inflation and rising interest rates, the path ahead for defense companies may be particularly complicated. Most of their contracts specify that the DOD will pay a fixed price for goods and services over a relatively long period, so they cannot compensate for higher costs by increasing prices. As defense companies begin to formulate a response to inflation, they will benefit from a strong understanding of their contract portfolio, including how common terms may impose constraints when developing a strategy to deal with rising prices. This understanding will help them work with the DOD to maximize the nation’s defense systems. ... '