In late March, Polestar announced that the single-motor Long Range variant of its Polestar 2 EV coupe would be arriving imminently upon US shores and be available starting at $45,900 — $33,400 after federal and state incentives — while its dual-motor sibling would start at $51,200. On Wednesday, the EV automaker announced that those prices would be going up. The single-motor variant will now start at $48,400 — $40,900 after the $7,500 federal tax credit — while the dual-motor AWD version will set buyers back $51,900 ($44,400 after the credit).
The price hike is due in part to the new standard features, updates and upgrades applied to the platform over the past 6 weeks, according to a Polestar spokesperson. For those extra few hundreds to thousands of dollars, PS2 buyers will have access not only to the hundred-plus OTA software updates that have already been released — including the one that boosts the dual-motor’s driving range to a respectable 260 miles — but the new high-efficiency heat pump announced in April and a more sensitive air quality sensor as well. That air quality sensor is part of the $4,200 Plus Pack and can show the driver “a breakdown of the air circulating outside of the vehicle, including pollen types,” according to the release. Similarly, ordering the Performance Pack (a mere $5,500) will include the recent software upgrade that squeezes an extra 68 HP and 15 lb-ft of torque out of the dual motors.
PS2 shoppers will also have their pick of two new exterior color options — a metallic shade called “Jupiter” and the same metallic black “Space” found on the PS1 — and a light grey “Zinc” option for the interior Nappa leather. Both the 19- and 20-inch rims designs have been updated too.
After decades on the decline intro, America’s labor movement is undergoing a massive renaissance with Starbucks, Amazon and Apple Store employees leading the way. Though the tech sector has only just begun basking in the newfound glow of collective bargaining rights, the automotive industry has a long been a hotbed for unionization. But the movement is not at all monolithic. In the excerpt below from her new book, Fight Like Hell: The Untold History of American Labor, journalist Kim Kelly recalls the summer of 1968 that saw the emergence of a new, more vocal UAW faction, the Dodge Revolutionary Union Movement, coincide with a flurry of wildcat strikes in Big Three plants across the Rust Belt.
As of 2021, the U.S. construction industry is still booming and the building trades are heavily unionized, but not all of the nation’s builders have been so lucky. The country’s manufacturing sector has declined severely since its post–World War II high point, and so has its union density. The auto industry’s shuttered factories and former jobs shipped to countries with lower wages and weaker unions have become a symbol of the waning American empire. But things weren’t always this dire. Unions once fought tooth and nail to establish a foothold in the country’s automobile plants, factories, and steel mills. When those workers were able to harness the power of collective bargaining, wages went up and working conditions improved. The American Dream, or at least, a stable middle class existence, became an achievable goal for workers without college degrees or privileged backgrounds. Many more became financially secure enough to actually purchase the products they made, boosting the economy as well as their sense of pride in their work. Those jobs were still difficult and demanding and carried physical risks, but those workers—or at least, some of those workers—could count on the union to have their back when injustice or calamity befell them.
In Detroit, those toiling on the assembly lines of the Big Three automakers—Chrysler, Ford, and General Motors—could turn to the United Auto Workers (UAW), then hailed as perhaps the most progressive “major” union in the country as it forced its way into the automotive factories of the mid-twentieth century. The UAW stood out like a sore thumb among the country’s many more conservative (and lily-white) unions, with leadership from the likes of former socialist and advocate of industrial democracy Walter Reuther and a strong history of support for the Civil Rights Movement. But to be clear, there was still much work to be done; Black representation in UAW leadership remained scarce despite its membership reaching nearly 30 percent Black in the late 1960s.
The Big Three had hired a wave of Black workers to fill their empty assembly lines during World War II, often subjecting them to the dirtiest and most dangerous tasks available and on-the-job racial discrimination. And then, of course, once white soldiers returned home and a recession set in, those same workers were the first ones sacrificed. Production picked back up in the 1960s, and Black workers were hired in large numbers once again. They grew to become a majority of the workforce in Detroit’s auto plants, but found themselves confronting the same problems as before. In factories where the union and the company had become accustomed to dealing with one another without much fuss, a culture of complacency set in and some workers began to feel that the union was more interested in keeping peace with the bosses than in fighting for its most vulnerable members. Tensions were rising, both in the factories and the world at large. By May 1968, as the struggle for Black liberation consumed the country, the memory of the 1967 Detroit riots remained fresh, and the streets of Paris were paralyzed by general strikes, a cadre of class-conscious Black activists and autoworkers saw an opportunity to press the union into action.
They called themselves DRUM—the Dodge Revolutionary Union Movement. DRUM was founded in the wake of a wildcat strike at Dodge’s Detroit plant, staffed by a handful of Black revolutionaries from the Black-owned, anti-capitalist Inner City Voice alternative newspaper. The ICV sprang up during the 1967 Detroit riots, published with a focus on Marxist thought and the Black liberation struggle. DRUM members boasted experience with other prominent movement groups like the Student Nonviolent Coordinating Committee and the Black Panthers, combining tactical knowledge with a revolutionary zeal attuned to their time and community.
General Gordon Baker, a seasoned activist and assembly worker at Chrysler’s Dodge Main plant, started DRUM with a series of clandestine meetings throughout the first half of 1968. By May 2, the group had grown powerful enough to see four thousand workers walk out of Dodge Main in a wildcat strike to protest the “speed-up” conditions in the plant, which saw workers forced to produce dangerous speed and work overtime to meet impossible quotas. Over the course of just one week, the plant had increased its output 39 percent. Black workers, joined by a group of older Polish women who worked in the plant’s trim shop, shut down the plant for the day, and soon bore the brunt of management’s wrath. Of the seven workers who were fired after the strike, five were Black. Among them was Baker, who sent a searing letter to the company in response to his dismissal. “In this day and age under the brutal repression reaped from the backs of Black workers, the leadership of a wildcat strike is a badge of honor and courage,” he wrote. “You have made the decision to do battle, and that is the only decision you will make. We shall decide the arena and the time.”
DRUM led another thousands-strong wildcat strike on July 8, this time shutting down the plant for two days and drawing in a number of Arab and white workers as well. Prior to the strike, the group had printed leaflets and held rallies that attracted hundreds of workers, students, and community members, a strategy DRUM would go on to use liberally in later campaigns to gin up support and spread its revolutionary message.
Men like Baker, Kenneth Cockrel, and Mike Hamlin were the public face of DRUM, but their work would have been impossible without the work of their female comrades, whose contributions were often overlooked. Hamlin admitted as much in his book-length conversation with longtime political activist and artist Michele Gibbs, A Black Revolutionary’s Life in Labor. “Possibly my deepest regret,” Hamlin writes, “is that we could not curb, much less transform, the doggish behavior and chauvinist attitudes of many of the men.”
Black women in the movement persevered despite this discrimination and disrespect at work, and they also found allies in unexpected places. Grace Lee Boggs, a Chinese American Marxist philosopher and activist with a PhD from Bryn Mawr, met her future husband James Boggs in Detroit after moving there in 1953. She and James, a Black activist, author (1963’s The American Revolution: Pages from a Negro Worker’s Notebook), and Chrysler autoworker, became fixtures in Detroit’s Black radical circles. They naturally fell in with the DRUM cadre, and Grace fit perfectly when Hamlin organized a DRUM-sponsored book club discussion forum in order to draw in progressive white and more moderate Black sympathizers. Interest in the Marxist book club was unexpectedly robust, and it grew to more than eight hundred members in its first year. Grace stepped in to help lead its discussion groups, and allowed young activists to visit her and James at their apartment and talk through thorny philosophical and political questions until the wee hours. She would go on to become one of the nation’s most respected Marxist political intellectuals and a lifelong activist for workers’ rights, feminism, Black liberation, and Asian American issues. As she told an interviewer prior to her death in 2015 at the age of one hundred, “People who recognize that the world is always being created anew, and we’re the ones that have to do it — they make revolutions.”
Further inside the DRUM orbit, Helen Jones, a printer, was the force behind the creation and distribution of their leaflets and publications. Women like Paula Hankins, Rachel Bishop, and Edna Ewell Watson, a nurse and confidant of Marxist scholar and former Black Panther Angela Davis, undertook their own labor organizing projects. In one case, the trio led a union drive among local hospital workers in the DRUM faction, hoping to carve out a place for female leadership within their movement. But ultimately, these expansion plans were dropped due to a lack of full support within DRUM. “Many of the male leaders acted as if women were sexual commodities, mindless, emotionally unstable, or invisible,” Edna Watson later told Dan Georgakas and Marvin Surkin for their Detroit: I Do Mind Dying. She claimed the organization held a traditionalist Black patriarchal view of women, in which they were expected to center and support their male counterparts’ needs at the expense of their own agenda. “There was no lack of roles for women… as long as they accepted subordination and invisibility.”
By 1969, the movement had spread to multiple other plants in the city, birthing groups like ELRUM (Eldon Avenue RUM), JARUM (Jefferson Avenue RUM), and outliers like UPRUM (UPS workers) and HRUM (healthcare workers). The disparate RUM groups then combined forces, forming the League of Revolutionary Black Workers. The new organization was to be led by the principles of Marxism, Leninism, and Maoism, but the league was never an ideological monolith. Its seven-member executive committee could not fully cohere the different political tendencies of its board or its eighty-member deep inner control group. Most urgently, opinions diverged on what shape, if any, further growth should take.
The embattled game company announced via Twitter on Thursday that it will host a livestream premiere event Tuesday, May 3rd at 10 am Pacific on Reveal.Blizzard.com. This isn’t the first time that a console franchise has expanded into mobile — Call of Duty and Fortnite have already launched their own iterations for phones and tablets. There are precious few details as to what the game will entail (beyond being set in the Warcraft Universe) or what gameplay mechanics will be used so be sure to join us next Tuesday for more coverage of WoW’s newest foray into the realm of handhelds.
At the start of the year, Google announced the Privacy Sandbox on Android project, a new system designed to eventually replace today’s existing third-party cookie schemes and reinvent a more privacy-centered method for serving advertisements. After an initial round of alpha testing and feedback, Google announced on Thursday that the first developer’s preview of the sandbox is now available as part of Android 13 beta 1.
The Privacy Sandbox is a multi-year development effort that will “limit sharing of user data with third parties and operate without cross-app identifiers, including advertising ID,” Google wrote in a February announcement. “We’re also exploring technologies that reduce the potential for covert data collection, including safer ways for apps to integrate with advertising SDKs.”
This preview provides developers with early looks at the sandbox’s SDK Runtime and Topics API so that they can better understand how they’ll fit into their apps and processes once it is officially released. We first saw Topics API back in January. It pulls data from the Chrome browser to identify the user’s top five interests for the week, based on their search and browsing history. Those topics are then compared against a database of topics from the Interactive Advertising Bureau and Google’s own data. Partner publishers can then ping the Topics API, see what the user is currently into, and then serve the most appropriate ads without having to know every nitty-gritty detail about their potential customer.
Developers will also have access to an early version of the Fledge API. This allows sites to run “remarket” to existing users — ie, serving users ads to remind them that they left items in their shopping cart and should just check out already. The Sandbox comes with everything that developers will need to test it, including the Android SDK and 64-bit Android Emulator. The company intends to further refine the toolset over the coming months and welcomes feedback and questions from the developer community
They may not be able to shout “Eureka!” like their human colleagues but AI/ML system have shown immense potential in the field of compound discovery — whether that’s sifting through reams of data to find new therapeutic compounds or imagining new recipes using the ingredients’ flavor profiles. Now a team from Meta AI, working with researchers at the University of Illinois, Urbana-Champaign, have created an AI that can devise and refine formulas for increasingly high-strength, low-carbon concrete.
Traditional methods for creating concrete, of which we produce billions of tons every year, are far from ecologically friendly. In fact, they generate an estimated 8 percent of the annual global carbon dioxide emission total. Advances have been made in recent years to reduce the concrete industry’s carbon footprint (as well as in make the material more rugged, more resilient and even capable of charging EVs) but overall its production remains among the most carbon intensive in modern construction.
Reducing the amount of carbon that goes into concrete could be as simple as changing the ingredients that go into concrete. The material is made from four basic components: cement, aggregate, water and admixture (which act as doping agents). Cement is far and away the most carbon-intensive ingredient of the four so research has been made into reducing the amount of cement needed by supplementing it with lower-carbon materials like fly ash, slag, or ground glass.
Similarly, aggregate materials like gravel, crushed stone, sand might be replaced with recycled concrete. The problem is that there are dozens of potential ingredient materials that could be used and the ratio of their amounts all interact to influence the structural profile of the resulting concrete. In short, there are a whole slew of possible combinations for researchers to test, select, and refine; and working through those myriad options sequentially, at human speed, is going to take forever. So the Meta folks trained an AI to do it, much faster.
Working with Prof. Lav Varshney, electrical and computer engineering department, and Prof. Nishant Garg, civil engineering department, both of the University of Illinois at Urbana-Champaign, the team first trained the model using the Concrete Compressive Strength data set. This set includes more than 1,000 concrete formulas as well as their structural attributes, including seven-day and 28-day compressive strength data. The team determined the resulting concrete mixture’s carbon footprint using the Cement Sustainability Initiative’s Environmental Product Declaration (EPD) tool.
Of the generated list of potential formulas, the research team then selected the five most promising options and iteratively refined them until they met or exceeded the 7- and 28-day strength metrics while dropping carbon requirements by at least 40 percent. The refinement process took mere weeks and ended up generating a concrete formula that exceeded all of those requirements while replacing as much as 50 percent of the required cement with fly ash and slag. Meta then teamed with concrete company Ozinga, the folks who recently built Meta’s newest datacenter in Illinois, to further refine the formula and conduct real world testing.
Looking ahead, the Meta team hopes to further improve the formula’s 3- and 5-day strength profiles (basically ensuring it dries faster so the rest of the construction can move ahead sooner) and get a better understanding of how it cures under varying weather conditions like wind or high humidity.
Keeping an EV’s batteries within their optimum operating temperature range is essential to getting the highest performance and longest life cycles possible from them. Too hot and charge seeps from the cells, too cold and the vehicle’s range can drop up to 20 percent with charging sessions taking significantly longer than they would in warmer climes. This is why heat pumps, devices that scavenge waste heat from a vehicle’s engine components to provide power other systems, have been finding their ways into a number of electric autos in recent years. Tesla has added them to its Model Y, 3, and S Plaid; Polestar includes them with the PS2 single-motor, and Rivian, well, Rivian does it a little different, but on Monday, GM announced its latest entry into waste heat reclamation game with the debut of its “Ultium Energy Recovery” system.
The UER is “based around an advanced automotive grade heat pump that captures and repurposes otherwise wasted energy,” Tim Grewe, GM director of electrification strategy, said during a press call last week. “It’s more sophisticated than even the most advanced thermal heat pump that you would find in modern homes.”
“We could do several things with this energy,” he continued, “including increase the range of our EVs, power low-level electrical functions like heating, and even pre-conditioning of our battery for faster charging and acceleration.” For an EV like the new Hummer, the estimated ten percent increase in range that this system provides translates into an extra 30 miles of range. Similarly, this heat pump is what drives the Hummer’s Watts to Freedom launch control function, autonomously conditioning the battery temperatures to the optimal level with which to dump as much current they can, as fast as they can, in order to propel the 9,000-pound EV SUV from 0 to 60 in 3 seconds flat.
“It’s one of those situations where you want to get the magnets in the motor as cold as you can to give you ultimate torque going forward,” GM energy recovery system project manager, Lawrence Zeer, said on the call. “And then you want to warm up the battery because the battery is give you a little more power when they’re warmed up.” Zeer also points out that given the immense size of these batteries — the Hummer’s is rumored to weigh more than 2,900 pounds — “it’s got a lot of heat capacity to it.” Conversely, the pump will also automatically precondition the batteries if the driver selects an upcoming charging station from the nav computer and can cool the cabin as easily as it warms it.
GM plans to include the recovery system across its electric vehicle lineup including the Hummer EV, the Lyric, and the upcoming Blazer EV. And since the recovery system is already standard throughout GM’s EV offerings, folks who’ve pre-ordered their Lyric and Hummers won’t have to turn around and head back to the dealership for a service installation.
As with most every other aspect of modern society, computerization, augmentation and automation have hyper-accelerated the pace at which wars are prosecuted — and who better to help reshape the US military into a 21st century fighting force than an entire industry centered on moving fast and breaking things? In his latest book, War Virtually: The Quest to Automate Conflict, Militarize Data, and Predict the Future, professor and chair of the Anthropology Department at San José State University, Roberto J González examines the military’s increasing reliance on remote weaponry and robotic systems are changing the way wars are waged. In the excerpt below, González investigates Big Tech’s role in the Pentagon’s high-tech transformations.
Ash Carter’s plan was simple but ambitious: to harness the best and brightest ideas from the tech industry for Pentagon use. Carter’s premise was that new commercial companies had surpassed the Defense Department’s ability to create cutting-edge technologies. The native Pennsylvanian, who had spent several years at Stanford University prior to his appointment as defense secretary, was deeply impressed with the innovative spirit of the Bay Area and its millionaire magnates. “They are inventing new technology, creating prosperity, connectivity, and freedom,” he said. “They feel they too are public servants, and they’d like to have somebody in Washington they can connect to.” Astonishingly, Carter was the first sitting defense secretary to visit Silicon Valley in more than twenty years.
The Pentagon has its own research and development agency, DARPA, but its projects tend to pursue objectives that are decades, not months, away. What the new defense secretary wanted was a nimble, streamlined office that could serve as a kind of broker, channeling tens or even hundreds of millions of dollars from the Defense Department’s massive budget toward up-and-coming firms developing technologies on the verge of completion. Ideally, DIUx would serve as a kind of liaison, negotiating the needs of grizzled four-star generals, the Pentagon’s civilian leaders, and hoodie-clad engineers and entrepreneurs. Within a year, DIUx opened branch offices in two other places with burgeoning tech sectors: Boston, Massachusetts, and Austin, Texas.
In the short term, Carter hoped that DIUx would build relationships with local start-ups, recruit top talent, get military reservists involved in projects, and streamline the Pentagon’s notoriously cumbersome procurement processes. “The key is to contract quickly — not to make these people fill out reams of paperwork,” he said. His long-term goals were even more ambitious: to take career military officers and assign them to work on futuristic projects in Silicon Valley for months at a time, to “expose them to new cultures and ideas they can take back to the Pentagon… [and] invite techies to spend time at Defense.”
In March 2016, Carter organized the Defense Innovation Board (DIB), an elite brain trust of civilians tasked with providing advice and recommendations to the Pentagon’s leadership. Carter appointed former Google CEO (and Alphabet board member) Eric Schmidt to chair the DIB, which includes current and former executives from Facebook, Google, and Instagram, among others.
Three years after Carter launched DIUx, it was renamed the Defense Innovation Unit (DIU), indicating that it was no longer experimental. This signaled the broad support the office had earned from Pentagon leaders. The Defense Department had lavished nearly $100 million on projects from forty-five companies, almost none of which were large defense contractors. Despite difficulties in the early stages — and speculation that the Trump administration might not support an initiative focused on regions that tended to skew toward the Democratic Party — DIUx was “a proven, valuable asset to the DoD,” in the words of Trump’s deputy defense secretary, Patrick Shanahan. “The organization itself is no longer an experiment,” he noted in an August 2018 memo, adding: “DIU remains vital to fostering innovation across the Department and transforming the way DoD builds a more lethal force.” Defense Secretary James “Mad Dog” Mattis visited Amazon’s Seattle headquarters and Google’s Palo Alto office in August 2017 and had nothing but praise for the tech industry. “I’m going out to see what we can pick up in DIUx,” he told reporters. In early 2018, the Trump administration requested a steep increase in DIU’s budget for fiscal year 2019, from $30 million to $71 million. For 2020, the administration requested $164 million, more than doubling the previous year’s request.
Q BRANCH
Although Pentagon officials portrayed DIUx as a groundbreaking organization, it was actually modeled after another firm established to serve the US Intelligence Community in a similar way. In the late 1990s, Ruth David, the CIA’s deputy director for science and technology, suggested that the agency needed to move in a radically new direction to ensure that it could capitalize on innovations being developed in the private sector, with a special focus on Silicon Valley firms. In 1999, under the leadership of its director, George Tenet, the CIA established a nonprofit legal entity called Peleus to fulfill this objective, with help from former Lockheed Martin CEO Norman Augustine. Soon after, the organization was renamed In-Q-Tel.
The first CEO, Gilman Louie, was an unconventional choice to head the enterprise. Louie had spent nearly twenty years as a video game developer who, among other things, created a popular series of Falcon F-16 flight simulators. At the time he agreed to join the new firm, he was chief creative officer for the toy company Hasbro. In a 2017 presentation at Stanford University, Louie claimed to have proposed that In-Q-Tel take the form of a venture capital fund. He also described how, at its core, the organization was created to solve “the big data problem”:
The problem they [CIA leaders] were trying to solve was: How to get technology companies who historically have never engaged with the federal government to actually provide technologies, particularly in the IT space, that the government can leverage. Because they were really afraid of what they called at that time the prospects of a “digital Pearl Harbor” Pearl Harbor
happened with every different part of the government having a piece of information but they couldn’t stitch it together to say, “Look, the attack at Pearl Harbor is imminent.” The White House had a piece of information, naval intelligence had a piece of information, ambassadors had a piece of information, the State Department had a piece of information, but they couldn’t put it all together [In] 1998, they began to realize that information was siloed across all these different intelligence agencies of which they could never stitch it together [F]undamentally what they were trying to solve was the big data problem. How do you stitch that together to get intelligence out of that data?
Louie served as In-Q-Tel’s chief executive for nearly seven years and played a crucial role in shaping the organization.
By channeling funds from intelligence agencies to nascent firms building technologies that might be useful for surveillance, intelligence gathering, data analysis, cyberwarfare, and cybersecurity, the CIA hoped to get an edge over its global rivals by using investment funds to co-opt creative engineers, hackers, scientists, and programmers. The Washington Post reported that “In-Q-Tel was engineered with a bundle of contradictions built in. It is independent of the CIA, yet answers wholly to it. It is a non- profit, yet its employees can profit, sometimes handsomely, from its work. It functions in public, but its products are strictly secret.” In 2005, the CIA pumped approximately $37 million into In-Q-Tel. By 2014, the organization’s funding had grown to nearly $94 million a year and it had made 325 investments with an astonishing range of technology firms, almost none of which were major defense contractors.
If In-Q-Tel sounds like something out of a James Bond movie, that’s because the organization was partly inspired by — and named after — Q Branch, a fictional research and development office of the British secret service, popularized in Ian Fleming’s spy novels and in the Hollywood blockbusters based on them, going back to the early 1960s. Ostensibly, both In-Q-Tel and DIUx were created to transfer emergent private-sector technologies into the US intelligence and military agencies, respectively. A somewhat different interpretation is that these organizations were launched “to capture technological innovations… [and] to capture new ideas.” From the perspective of the CIA these arrangements have been a “win-win,” but critics have described them as a boondoggle — lack of transparency, oversight, and streamlined procurement means that there is great potential for conflicts of interest. Other critics point to In-Q-Tel as a prime example of the militarization of the tech industry.
There’s an important difference between DIUx and In-Q-Tel. DIUx is part of the Defense Department and is therefore financially dependent on Pentagon funds. By contrast, In-Q-Tel is, in legal and financial terms, a distinct entity. When it invests in promising companies, In-Q-Tel also becomes part owner of those firms. In monetary and technological terms, it’s likely that the most profitable In-Q-Tel investment was funding for Keyhole, a San Francisco–based company that developed software capable of weaving together satellite images and aerial photos to create three-dimensional models of Earth’s surface. The program was capable of creating a virtual high-resolution map of the entire planet. In-Q-Tel provided funding in 2003, and within months, the US military was using the software to support American troops in Iraq.
Official sources never revealed how much In-Q-Tel invested in Keyhole. In 2004, Google purchased the start-up for an undisclosed amount and renamed it Google Earth. The acquisition was significant. Yasha Levine writes that the Keyhole-Google deal “marked the moment the company stopped being a purely consumer-facing internet company and began integrating with the US government [From Keyhole, Google] also acquired an In-Q-Tel executive named Rob Painter, who came with deep connections to the world of intelligence and military contracting.” By 2006 and 2007, Google was actively seeking government contracts “evenly spread among military, intelligence, and civilian agencies,” according to the Washington Post.
Apart from Google, several other large technology firms have acquired startups funded by In-Q-Tel, including IBM, which purchased the data storage company Cleversafe; Cisco Systems, which absorbed a conversational AI interface startup called MindMeld; Samsung, which snagged nanotechnology display firm QD Vision; and Amazon, which bought multiscreen video delivery company Elemental Technologies. While these investments have funded relatively mundane technologies, In-Q-Tel’s portfolio includes firms with futuristic projects such as Cyphy, which manufactures tethered drones that can fly reconnaissance missions for extended periods, thanks to a continuous power source; Atlas Wearables, which produces smart fitness trackers that closely monitor body movements and vital signs; Fuel3d, which sells a handheld device that instantly produces detailed three-dimensional scans of structures or other objects; and Sonitus, which has developed a wireless communication system, part of which fits inside the user’s mouth. If DIUx has placed its bets with robotics and AI companies, In-Q-Tel has been particularly interested in those creating surveillance technologies — geospatial satellite firms, advanced sensors, biometrics equipment, DNA analyzers, language translation devices, and cyber-defense systems.
More recently, In-Q-Tel has shifted toward firms specializing in data mining social media and other internet platforms. These include Dataminr, which streams Twitter data to spot trends and potential threats; Geofeedia, which collects geographically indexed social media messages related to breaking news events such as protests; PATHAR, a company specializing in social network analysis; and TransVoyant, a data integration firm that collates data from satellites, radar, drones, and other sensors. In-Q-Tel has also created Lab41, a Silicon Valley technology center specializing in big data analysis and machine learning.
Big box electronics retailer Best Buy announced the launch of a new appliance recycling program Thursday that will allow customers to have up to two large pieces (and an unlimited amount of small items) of unwanted tech hauled away for a $200 fee.
Best Buy already operates a number of consumer electronics recycling programs, with customers either dropping off items in-store for gift cards or paying a $30 to 50 fee to have their old appliances taken away when their Best Buy-bought replacements are delivered. This new Standalone Haul-Away service does not require any additional purchases and comes with a 20 percent discount to Best Buy TotalTech subscribers (a $200-a-year scheme that includes on-demand GeekSquad access). However Haul-Away is limited in what it can take.
Customers can get rid of up to two (2) all-in-one computers, TVs, large appliances and refrigerators as well as as many hard drives, gaming consoles, laptops and un-regiftable immersion blenders as they are willing to part with. If you’re looking to offload old musical instruments, DVDs, software, or legacy formats, on the other hand, you’d best look elsewhere because Best Buy won’t take them.
On November 21st, Trevor Jacob’s single-engine airplane fell out of the sky — a harrowing experience that the YouTuber just so happened to catch on film and upload to social media. In January, aviation experts began investigating the incident (as they are wont to do in the event of most every aviation crash) and, on Thursday, the Federal Aviation Administration formally accused Jacob of staging the entire incident and intentionally crashing his 1940 Taylorcraft for online clout.
At the time, Jacob, a former Olympic snowboarder, claimed that his plane had malfunctioned, forcing him to bail out and parachute to safely while the aircraft crashed into the Los Padres National Forest in Southern California. However, in a letter dated April 11th, the FAA informed him that he had operated his plane in a “careless or reckless manner so as to endanger the life or property of another,” a violation of aviation regulations. The FAA also revoked his pilot’s license effective immediately.
When reached by the New York Times this week, Jacob claimed to not be aware of the April 11th letter but declined to comment, on advice of his attorney. Although the FAA can’t actually prosecute anybody for violating regulations, should Jacob fail to surrender his pilot’s license he can be held liable for “further legal enforcement action” and fined up to $1,644 a day until he does.
For humans, identifying items in a scene — whether that’s an avocado or an Aventador, a pile of mashed potatoes or an alien mothership — is as simple as looking at them. But for artificial intelligence and computer vision systems, developing a high-fidelity understanding of their surroundings takes a bit more effort. Well, a lot more effort. Around 800 hours of hand-labeling training images effort, if we’re being specific. To help machines better see the way people do, a team of researchers at MIT CSAIL in collaboration with Cornell University and Microsoft have developed STEGO, an algorithm able to identify images down to the individual pixel.
MIT CSAIL
Normally, creating CV training data involves a human drawing boxes around specific objects within an image — say, a box around the dog sitting in a field of grass — and labeling those boxes with what’s inside (“dog”), so that the AI trained on it will be able to tell the dog from the grass. STEGO (Self-supervised Transformer with Energy-based Graph Optimization), conversely, uses a technique known as semantic segmentation, which applies a class label to each pixel in the image to give the AI a more accurate view of the world around it.
Whereas a labeled box would have the object plus other items in the surrounding pixels within the boxed-in boundary, semantic segmentation labels every pixel in the object, but only the pixels that comprise the object — you get just dog pixels, not dog pixels plus some grass too. It’s the machine learning equivalent of using the Smart Lasso in Photoshop versus the Rectangular Marquee tool.
The problem with this technique is one of scope. Conventional multi-shot supervised systems often demand thousands, if not hundreds of thousands, of labeled images with which to train the algorithm. Multiply that by the 65,536 individual pixels that make up even a single 256×256 image, all of which now need to be individually labeled as well, and the workload required quickly spirals into impossibility.
Instead, “STEGO looks for similar objects that appear throughout a dataset,” the CSAIL team wrote in a press release Thursday. “It then associates these similar objects together to construct a consistent view of the world across all of the images it learns from.”
“If you’re looking at oncological scans, the surface of planets, or high-resolution biological images, it’s hard to know what objects to look for without expert knowledge. In emerging domains, sometimes even human experts don’t know what the right objects should be,” MIT CSAIL PhD student, Microsoft Software Engineer, and the paper’s lead author Mark Hamilton said. “In these types of situations where you want to design a method to operate at the boundaries of science, you can’t rely on humans to figure it out before machines do.”
Trained on a wide variety of image domains — from home interiors to high altitude aerial shots — STEGO doubled the performance of previous semantic segmentation schemes, closely aligning with the image appraisals of the human control. What’s more, “when applied to driverless car datasets, STEGO successfully segmented out roads, people, and street signs with much higher resolution and granularity than previous systems. On images from space, the system broke down every single square foot of the surface of the Earth into roads, vegetation, and buildings,” the MIT CSAIL team wrote.
MIT CSAIL
“In making a general tool for understanding potentially complicated data sets, we hope that this type of an algorithm can automate the scientific process of object discovery from images,” Hamilton said. “There’s a lot of different domains where human labeling would be prohibitively expensive, or humans simply don’t even know the specific structure, like in certain biological and astrophysical domains. We hope that future work enables application to a very broad scope of data sets. Since you don’t need any human labels, we can now start to apply ML tools more broadly.”
Despite its superior performance to the systems that came before it, STEGO does have limitations. For example, it can identify both pasta and grits as “food-stuffs” but doesn’t differentiate between them very well. It also gets confused by nonsensical images, such as a banana sitting on a phone receiver. Is this a food-stuff? Is this a pigeon? STEGO can’t tell. The team hopes to build a bit more flexibility into future iterations, allowing the system to identify objects under multiple classes.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/10835833/n4scgse21iuz.jpg){kind=link}