According to Bloomberg, China's chip import value dropped significantly by 15.4% in 2023, from $413 billion to $349 billion. "Chip sales were down across the board in 2023 thanks to a weakening global economy, but China's chip imports indicate that its economy might be in trouble," reports Tom's Hardware. "The country's inability to import cutting-edge silicon is also certainly a factor in its decreasing chip imports." From the report: In 2022, the value of chip imports to China stood at $413 billion, and in 2023 the country only imported chips worth a total of $349 billion, a 15.4% decrease in value. That a drop happened at all isn't surprising; even TSMC, usually considered to be one of the most advanced fabbing corporation in the world, saw its sales decline by 4.5%. However, a 15.4% decrease in shipments is much more significant, and indicates China has particular issues other than weaker demand across the world.

China's ongoing economic issues, such as its high deflation could play a part. Deflation is when currency increases in value, the polar opposite of inflation, when currency loses value. As inflation has been a significant problem for countries such as the U.S. and UK, deflation might sound much more appealing, but economically it can be problematic. A deflationary economy encourages consumers not to spend, since money is increasing in value, meaning buyers can purchase more if they wait. In other words, deflation decreases demand for products like semiconductors.

However, shipment volume only decreased by 10.8% compared to the 15.4% decline in value, meaning the chips that China didn't buy in 2023 were particularly valuable. This likely reflects U.S. sanctions on China, which prevents it from buying top-end graphics cards, especially from Nvidia. The H100, H200, GH200, and the RTX 4090 are illegal to ship to China, and they're some of Nvidia's best GPUs. The moving target for U.S. sanctions could also make exporters and importers more tepid, as it's hard to tell if more sanctions could suddenly upend plans and business deals.

An anonymous reader quotes a report from Electrek: The world's first floating offshore wind farm, Hywind Scotland, is coming offline for three to four months for "heavy maintenance." Hywind Scotland's operator, Norwegian power giant Equinor, says that operational data has indicated that its wind turbines need work. The pilot project has been in operation since 2017. The five Siemens Gamesa turbines will be towed to Norway this summer. An Equinor spokesperson said, "This is the first such operation for a floating farm, and the safest method to do this is to tow the turbines to shore and execute the operations in sheltered conditions."

Norwegian contractor Wergeland Group will undertake the work. The spokesperson added, "Wergeland is the closest port with offshore wind experience and sufficient water depth that can service these turbines." As the world's first floating offshore wind farm, Hywind Scotland has trailblazed for much larger floating wind farms now in the pipeline. Its five floating wind turbines have a total capacity of 30 megawatts (MW). It generates enough electricity to power the equivalent of 34,000 households in the UK. Each turbine's maximum height, base to turbine, is 253 meters (830 feet). [...] Equinor said in December 2022, when Hywind Scotland turned five, that it was the world's best-performing offshore wind farm, achieving a capacity factor of 54% over its five years of operations.

America's Environmental Protection Agency "has signed off on a California oil company's plans to permanently store carbon emissions deep underground to combat global warming," reports the Los Angeles Times: California Resources Corp., the state's largest oil and gas company, applied for permission to send 1.46 million metric tons of carbon dioxide each year into the Elk Hills oil field, a depleted oil reservoir about 25 miles outside of downtown Bakersfield. The emissions would be collected from several industrial sources nearby, compressed into a liquid-like state and injected into porous rock more than one mile underground.

Although this technique has never been performed on a large scale in California, the state's climate plan calls for these operations to be widely deployed across the Central Valley to reduce carbon emissions from industrial facilities. The EPA issued a draft permit for the California Resources Corp. project, which is poised to be finalized in March following public comments. As California transitions away from oil production, a new business model for fossil fuel companies has emerged: carbon management. Oil companies have heavily invested in transforming their vast network of exhausted oil reservoirs into a long-term storage sites for planet-warming gases, including California Resources Corp., the largest nongovernmental owner of mineral rights in California...

[Environmentalists] say that the transportation and injection of CO2 — an asphyxiating gas that displaces oxygen — could lead to dangerous leaks. Nationwide, there have been at least 25 carbon dioxide pipeline leaks between 2002 and 2021, according to the U.S. Department of Transportation. Perhaps the most notable incident occurred in Satartia, Miss., in 2020 when a CO2 pipeline ruptured following heavy rains. The leak led to the hospitalization of 45 people and the evacuation of 200 residents... Under the EPA draft permit, California Resources Corp. must take a number of steps to mitigate these risks. The company must plug 157 wells to ensure the CO2 remains underground, monitor the injection site for leaks and obtain a $33-million insurance policy.
Canadian-based Brookfield Corporation also invested $500 million, according to the article, with California Resources Corp. seeking permits for five projects — more than any company in the nation. "It's kind of reversing the role, if you will," says their chief sustainability officer. "Instead of taking oil and gas out, we're putting carbon in."

Meanwhile, there's applications for "about a dozen" more projects in California's Central Valley that could store millions of tons of carbon emissions in old oil and gas fields — and California Resources Corp says greater Los Angeles is also "being evaluated" as a potential storage site.

Generative AI "doesn't easily carry over into robotics," write two researchers in IEEE Spectrum, "because the Internet is not full of robotic-interaction data in the same way that it's full of text and images."

That's why they're working on a single deep neural network capable of piloting many different types of robots... Robots need robot data to learn from, and this data is typically created slowly and tediously by researchers in laboratory environments for very specific tasks... The most impressive results typically only work in a single laboratory, on a single robot, and often involve only a handful of behaviors... [W]hat if we were to pool together the experiences of many robots, so a new robot could learn from all of them at once? We decided to give it a try. In 2023, our labs at Google and the University of California, Berkeley came together with 32 other robotics laboratories in North America, Europe, and Asia to undertake the RT-X project, with the goal of assembling data, resources, and code to make general-purpose robots a reality...

The question is whether a deep neural network trained on data from a sufficiently large number of different robots can learn to "drive" all of them — even robots with very different appearances, physical properties, and capabilities. If so, this approach could potentially unlock the power of large datasets for robotic learning. The scale of this project is very large because it has to be. The RT-X dataset currently contains nearly a million robotic trials for 22 types of robots, including many of the most commonly used robotic arms on the market...

Surprisingly, we found that our multirobot data could be used with relatively simple machine-learning methods, provided that we follow the recipe of using large neural-network models with large datasets. Leveraging the same kinds of models used in current LLMs like ChatGPT, we were able to train robot-control algorithms that do not require any special features for cross-embodiment. Much like a person can drive a car or ride a bicycle using the same brain, a model trained on the RT-X dataset can simply recognize what kind of robot it's controlling from what it sees in the robot's own camera observations. If the robot's camera sees a UR10 industrial arm, the model sends commands appropriate to a UR10. If the model instead sees a low-cost WidowX hobbyist arm, the model moves it accordingly.
"To test the capabilities of our model, five of the laboratories involved in the RT-X collaboration each tested it in a head-to-head comparison against the best control system they had developed independently for their own robot... Remarkably, the single unified model provided improved performance over each laboratory's own best method, succeeding at the tasks about 50 percent more often on average." And they then used a pre-existing vision-language model to successfully add the ability to output robot actions in response to image-based prompts.

"The RT-X project shows what is possible when the robot-learning community acts together... and we hope that RT-X will grow into a collaborative effort to develop data standards, reusable models, and new techniques and algorithms."

Thanks to long-time Slashdot reader Futurepower(R) for sharing the article.

"Chinese company Betavolt has announced an atomic energy battery for consumers with a touted 50-year lifespan," reports Tom's Hardware: The Betavolt BV100 will be the first product to launch using the firm's new atomic battery technology, constructed using a nickel -63 isotope and diamond semiconductor material. Betavolt says that its nuclear battery will target aerospace, AI devices, medical, MEMS systems, intelligent sensors, small drones, and robots — and may eventually mean manufacturers can sell smartphones that never need charging...

[T]he BV100, which is in the pilot stage ahead of mass production, doesn't offer a lot of power. This 15 x 15 x 5mm battery delivers 100 microwatts at 3 volts. It is mentioned that multiple BV100 batteries can be used together in series or parallel depending on device requirements. Betavolt also asserts that it has plans to launch a 1-watt version of its atomic battery in 2025. The new BV100 is claimed to be a disruptive product on two counts. Firstly, a safe miniature atomic battery with 50 years of maintenance-free stamina is a breakthrough. Secondly, Betavolt claims it is the only company in the world with the technology to dope large-size diamond semiconductor materials, as used by the BV100. It is using its 4th Gen diamond semiconductor material here...

[T]he Betavolt BV100 is claimed to be safe for consumers and won't leak radiation even if subjected to gunshots or puncture... Betavolt's battery uses a nickel -63 isotope as the energy source, which decays to a stable isotope of copper. This, plus the diamond semiconductor material, helps the BV100 operate stably in environments ranging from -60 to 120 degrees Celsius, according to the firm...

Betavolt will be well aware of devices with a greater thirst for power and teases that it is investigating isotopes such as strontium- 90, promethium- 147, and deuterium to develop atomic energy batteries with higher power levels and even longer service lives — up to 230 years.
Thanks to long-time Slashdot reader hackingbear for sharing the news.

A new analysis suggests that wind turbines have little impact on bird populations, according to the Economist — and that oil-and-gas extraction may be worse: Erik Katovich [an economist at the University of Geneva] combined bird population and species maps with the locations and construction dates of all wind turbines in the United States, with the exceptions of Alaska and Hawaii, between 2000 and 2020. He found that building turbines had no discernible effect on bird populations. That reassuring finding held even when he looked specifically at large birds like hawks, vultures and eagles that many people believe are particularly vulnerable to being struck.

But Dr. Katovich did not confine his analysis to wind power alone. He also examined oil-and-gas extraction. Like wind power, this has boomed in America over the past couple of decades, with the rise of shale gas produced by hydraulic fracturing, or fracking, of rocks. Production has risen from 37m cubic metres in 2007 to 740m cubic metres in 2020. Comparing bird populations to the locations of new gas wells revealed an average 15% drop in bird numbers when new wells were drilled, probably due to a combination of noise, air pollution and the disturbance of rivers and ponds that many birds rely upon. When drilling happens in places designated by the National Audubon Society as "important bird areas", bird numbers instead dropped by 25%. Such places are typically migration hubs, feeding grounds or breeding locations.

Wind power, in other words, not only produces far less planet-heating carbon dioxide and methane than do fossil fuels. It appears to be significantly less damaging to wildlife, too.
Thanks to long-time Slashdot reader SpzToid for sharing the article.

Friday America's Environmental Protection Agency "proposed steep new fees on methane emissions from oil and gas facilities," reports the Washington Post, "escalating a crackdown on the fossil fuel industry's planet-warming pollution."

Methane does not linger in the atmosphere as long as carbon dioxide, but it is far more effective at trapping heat — roughly 80 times more potent in its first decade. It is responsible for roughly a third of global warming today, and the oil and gas industry accounts for about 14 percent of the world's annual methane emissions, according to estimates from the International Energy Agency. Other large methane sources include livestock, landfills and coal mines.
So America's new Methane Emissions Reduction Program "levies a fee on wasteful methane emissions from large oil and gas facilities," according to the article: The fee starts at $900 per metric ton of emissions in 2024, increasing to $1,200 in 2025 and $1,500 in 2026 and thereafter. The EPA proposal lays out how the fee will be implemented, including how the charge will be calculated...

At the U.N. Climate Change Conference in Dubai in December, EPA Administrator Michael Regan announced final standards to limit methane emissions from U.S. oil and gas operations. Fossil fuel companies that comply with these standards will be exempt from the new fee... Fred Krupp, president of the Environmental Defense Fund, said the fee will encourage fossil fuel firms to deploy innovative technologies that detect methane leaks. Such cutting-edge technologies range from ground-based sensors to satellites in space. "Proven solutions to cut oil and gas methane and to avoid the fee are being used by leading companies in states across the country," Krupp said in a statement...

In addition to methane, the EPA proposal could slash emissions of hazardous air pollutants, including smog-forming volatile organic compounds and cancer-causing benzene [according to an EPA official].
The federal government also gave America's fossil fuel companies nearly $1 billion to help them comply with the methane regulation, according to the article.

The article also includes this statement from an executive at the American Petroleum Institute, the top lobbying arm of the U.S. oil and gas industry, complaining that the fines create a "regime" that would "stifle innovation," and urging Congress to repeal it.

At CES 2024 this week, Micron demonstrated its next-gen LPCAMM2 memory modules based on LPDDR5X memory. Not only are they smaller and more powerful than traditional SODIMMs, they can be "serviced during the manufacturing process and upgraded by the user," says Micron. Tom's Hardware reports: Micron's LPCAMM2 are industry-standard memory modules that will be available in 16 GB, 32 GB, and 64 GB capacities as well as with speed bins of up to a 9600 MT/s data transfer rate. These modules are designed to replace conventional SODIMMs as well as soldered-down LPDDR5X memory subsystem while offering the best of both worlds: flexibility, repairability, and upgradeability of modular memory solutions as well as high performance and low power consumption of mobile DRAM. Indeed, a Micron LPCAMM2 module is smaller than a traditional SODIMM despite the fact that it has a 128-bit memory interface and up to 64 GB of LPDDR5X memory onboard. Needless to say, the module is massively smaller than two SODIMM memory sticks that offer a 128-bit memory interface both in terms of height and in terms of physical footprint.

PC and microchip companies struggling to get consumers to replace pandemic-era laptops offered a new feature to crowds this week at CES: AI. From a report: PC and chipmakers including AMD and Intel are betting that the so-called "neural processing units" now found in the latest chip designs will encourage consumers to once again pay for higher-end laptops. Adding additional AI capabilities could help take market share from Apple. "The conversations I'm having with customers are about 'how do I get my PC ready for what I think is coming in AI and going to be able to deliver,'" said Sam Burd, Dell Technologies' president of its PC business. Chipmakers built the NPU blocks because they can achieve a high level of performance for AI functions with relatively modest power needs. Today there are few applications that might take full advantage of the new capabilities, but more are coming, said David McAfee, corporate vice president and general manager of the client channel business at AMD.

Among the few applications that can take advantage of such chips is the creative suite of software produced by Adobe. Intel hosted an "open house" where a handful of PC vendors showed off their latest laptops with demos designed to put the new capabilities on display. Machines from the likes of Dell and Lenovo were arrayed inside one of the cavernous ballrooms at the Venetian Convention Center on Las Vegas Boulevard.

An anonymous reader quotes a report from The Guardian: Joe Biden's administration has unveiled $623 million in funding to boost the number of electric vehicle charging points in the U.S., amid concerns that the transition to zero-carbon transportation isn't keeping pace with goals to tackle the climate crisis. The funding will be distributed in grants for dozens of programs across 22 states, such as EV chargers for apartment blocks in New Jersey, rapid chargers in Oregon and hydrogen fuel chargers for freight trucks in Texas. In all, it's expected the money, drawn from the bipartisan infrastructure law, will add 7,500 chargers to the US total.

There are about 170,000 electric vehicle chargers in the U.S., a huge leap from a network that was barely visible prior to Biden taking office, and the White House has set a goal for 500,000 chargers to help support the shift away from gasoline and diesel cars. "The U.S. is taking the lead globally on electric vehicles," said Ali Zaidi, a climate adviser to Biden who said the US is on a trajectory to "meet and exceed" the administration's charger goal. "We will continue to see this buildout over the coming years and decades until we've achieved a fully net zero transportation sector," he added. On Thursday, the House approved legislation to undo a Biden administration rule meant to facilitate the proliferation of EV charging stations. "S. J. Res. 38 from Sen. Marco Rubio (R-Fla.), would scrap a Federal Highway Administration waiver from domestic sourcing requirements for EV chargers funded by the 2021 bipartisan infrastructure law. It already passed the Senate 50-48," reports Politico.

"A waiver undercuts domestic investments and risks empowering foreign nations," said Rep. Sam Graves (R-Mo.), chair of the Transportation and Infrastructure Committee, during House debate Thursday. "If the administration is going to continue to push for a massive transition to EVs, it should ensure and comply with Buy America requirements." The White House promised to veto it and said it would backfire, saying it was so poorly worded it would actually result in fewer new American-made charging stations.

Julian Spector reports via Canary Media: Hawaii shut down its last coal plant on September 1, 2022, eliminating 180 megawatts of fossil-fueled baseload power from the grid on Oahu -- a crucial step in the state's first-in-the-nation commitment to cease burning fossil fuels for electricity by 2045. But the move posed a question that's becoming increasingly urgent as clean energy surges across the United States: How do you maintain a reliable grid while switching from familiar fossil plants to a portfolio of small and large renewables that run off the vagaries of the weather? Now Hawaii has an answer: It's a gigantic battery, unlike the gigantic batteries that have been built before.

The Kapolei Energy Storage system actually began commercial operations before Christmas on the industrial west side of Oahu, according to Plus Power, the Houston-based firm that developed and owns the project. Now, Kapolei's 158 Tesla Megapacks are charging and discharging based on signals from utility Hawaiian Electric. The plant's 185 megawatts of instantaneous discharge capacity match what the old coal plant could inject into the grid, though the batteries react far more quickly, with a 250-millisecond response time. Instead of generating power, they absorb it from the grid, ideally when it's flush with renewable generation, and deliver that cheap, clean power back in the evening hours when it's desperately needed.

The construction process had its setbacks, as did the broader effort to replace the coal plant with a roster of large-scale clean energy projects. The Kapolei battery was initially intended to come online before the coal plant retired. Covid disrupted deliveries for the grid battery industry across the board, and Kapolei's remote location in the middle of the Pacific Ocean didn't make things easier. By summer 2021, Plus Power was hoping to complete Kapolei by the end of 2022, but it ended up taking another year. Even then, it has joined the grid before several of the other large solar and battery projects slated to replace the coal plant's production with clean power.

An anonymous reader quotes a report from Ars Technica: On Tuesday, the quantum computing startup Quera laid out a road map that will bring error correction to quantum computing in only two years and enable useful computations using it by 2026, years ahead of when IBM plans to offer the equivalent. Normally, this sort of thing should be dismissed as hype. Except the company is Quera, which is a spinoff of the Harvard University lab that demonstrated the ability to identify and manage errors using hardware that's similar in design to what Quera is building. Also notable: Quera uses the same type of qubit that a rival startup, Atom Computing, has already scaled up to over 1,000 qubits. So, while the announcement should be viewed cautiously -- several companies have promised rapid scaling and then failed to deliver -- there are some reasons it should be viewed seriously as well. [...]

As our earlier coverage described, the Harvard lab where the technology behind Quera's hardware was developed has already demonstrated a key step toward error correction. It created logical qubits from small collections of atoms, performed operations on them, and determined when errors occurred (those errors were not corrected in these experiments). But that work relied on operations that are relatively easy to perform with trapped atoms: two qubits were superimposed, and both were exposed to the same combination of laser lights, essentially performing the same manipulation on both simultaneously. Unfortunately, only a subset of the operations that are likely to be desired for a calculation can be done that way. So, the road map includes a demonstration of additional types of operations in 2024 and 2025. At the same time, the company plans to rapidly scale the number of qubits. Its goal for 2024 hasn't been settled on yet, but [Quera's Yuval Boger] indicated that the goal is unlikely to be much more than double the current 256. By 2025, however, the road map calls for over 3,000 qubits and over 10,000 a year later. This year's small step will add pressure to the need for progress in the ensuing years.

If things go according to plan, the 3,000-plus qubits of 2025 can be combined to produce 30 logical qubits, meaning about 100 physical qubits per logical one. This allows fairly robust error correction schemes and has undoubtedly been influenced by Quera's understanding of the error rate of its current atomic qubits. That's not enough to perform any algorithms that can't be simulated on today's hardware, but it would be more than sufficient to allow people to get experience with developing software using the technology. (The company will also release a logical qubit simulator to help here.) Quera will undoubtedly use this system to develop its error correction process -- Boger indicated that the company expected it would be transparent to the user. In other words, people running operations on Quera's hardware can submit jobs knowing that, while they're running, the system will be handling the error correction for them. Finally, the 2026 machine will enable up to 100 logical qubits, which is expected to be sufficient to perform useful calculations, such as the simulation of small molecules. More general-purpose quantum computing will need to wait for higher qubit counts still.

HP has used its "Dynamic Security" firmware updates to "create a monopoly" of replacement printer ink cartridges, a lawsuit filed against the company on January 5 claims. From a report: The lawsuit, which is seeking class-action certification, represents yet another form of litigation against HP for bricking printers when they try to use ink that doesn't bear an HP logo. The lawsuit (PDF), which was filed in US District Court in the Northern District of Illinois, names 11 plaintiffs and seeks an injunction against HP requiring the company to disable its printer firmware updates from preventing the use of non-HP branded ink. The lawsuit also seeks monetary damages greater than $5,000,000 and a trial by jury. [...] HP was wrong to issue a firmware update affecting printer functionality, and users were not notified that accepting firmware updates "could damage any features of the printer," the lawsuit says.

home-electro.com shares a report from Engadget: On Sunday night Samsung held its annual First Look event at CES 2024, where the company teased the world's first transparent MicroLED display. While there's still no word on how much it costs or when this tech will find its way into retail devices, Samsung showcased its transparent MicroLED display side-by-side next to transparent OLED and transparent LCD models to really highlight the differences between the tech. Compared to the others, not only was the MicroLED panel significantly brighter, it also featured a completely frameless design and a more transparent glass panel that made it easier to see objects behind it. LG also unveiled a similar piece of tech: the company's "first wireless transparent OLED TV." It's called the OLED T and supports 4K resolution and LG's wireless transmission tech for audio and video.

You can watch a demo of Samsung's transparent microLED screen on YouTube.

An anonymous reader quotes a report from Ars Technica: In April, Microsoft announced that it would stop selling Microsoft-branded computer peripherals. Today, Onward Brands announced that it's giving those discarded Microsoft-stamped gadgets a second life under new branding. Products like the Microsoft Ergonomic Keyboard will become Incase products with "Designed by Microsoft" branding. Beyond the computer accessories saying "Designed by Microsoft," they should be the same keyboards, mice, webcams, headsets, and speakers, Onward, Incase's parent company, said, per The Verge. Onward said its Incase brand will bring back 23 Microsoft-designed products in 2024 and hopes for availability to start in Q2. Incase also plans to launch an ergonomic keyboard that Microsoft designed but never released. Onward CEO Charlie Tebele told The Verge that there's "potential" for Incase to release even more designs Microsoft never let us see.

The return of Microsoft peripheral designs resurrects (albeit in a new form) a line of computer gear started in 1983 when Microsoft released its first mouse, the Microsoft Mouse. Neither Onward nor Microsoft shared the full terms of their licensing agreement, but Onward claims that Incase will leverage the same supply chain and manufacturing components that Microsoft did, The Verge noted. "Microsoft will still retain ownership of its designs, so it could potentially bring back classic mice or keyboards itself in the future or continue to renew its license to Incase," The Verge reported, pointing out that Onward isn't licensing every single one of Microsoft's computer peripherals. Some classics, like the Intellimouse or its modern iterations, for example, don't make the Incase reboot list. For its part, Microsoft is still "convicted on going under one single" Surface brand, Nancie Gaskill, general manager of Surface, told The Verge. Further reading: Microsoft Adding New Key To PC Keyboards For First Time Since 1994

Nvidia on Monday announced its new GeForce RTX 40 SUPER series GPUs, promising significant performance gains for gaming, creative workflows and artificial intelligence capabilities over previous models. The new lineup includes the GeForce RTX 4080 SUPER, RTX 4070 Ti SUPER and RTX 4070 SUPER GPUs. Nvidia said the chips deliver up to 52 shader teraflops, 121 ray tracing teraflops and 836 AI teraflops.

The top-of-the-line RTX 4080 SUPER model will go on sale starting Jan. 31 priced from $999, while the RTX 4070 Ti SUPER and RTX 4070 SUPER will hit shelves on Jan. 24 and Jan. 17 respectively, priced at $799 and $599. The company said the new GPUs can accelerate ray tracing visuals in games by up to 4 times with Nvidia's Deep Learning Super Sampling (DLSS) technology. DLSS uses AI to boost frame rates in games while maintaining image quality.

Compared to its predecessor, the RTX 4080 SUPER is 1.4 times faster at 4K gaming than Nvidia's previous top gaming GPU, the RTX 3080 Ti, without DLSS enabled, Nvidia said. With DLSS Frame Generation switched on, the performance gap widens to 2 times as fast. The new GPU lineup also promises significant gains in AI workloads often used by creative professionals, such as video generation and image upscaling, Nvidia said.

Long-time Slashdot reader xetdog shared this report from IEEE Spectrum: High in the Andes mountains where the borders of Argentina, Bolivia, and Chile intersect, white expanses of salt stretch for thousands of kilometers. Under these flats lie reservoirs of brine that contain upwards of 58% of the world's lithium. For decades, producers have extracted that lithium by pumping the water up to the surface and letting it evaporate until the lithium salts become concentrated enough to filter out. The process takes 12 to 18 months, leaving behind piles of waste containing other metals. It also evaporates nearly 2 million liters of local water resources, harming indigenous communities.

To keep up, many companies are now developing processes to chemically or physically filter out lithium from brines and inject the brine back underground. These direct lithium extraction (DLE) technologies take hours instead of months and could double the production of lithium from existing brine operations. Much as shale extraction did for oil, DLE is a "potential game-changing technology for lithium supply," because it could unlock new sources of lithium, according to a recent report by Goldman Sachs. But in contrast to shale's fracking risks, DLE brings environmental benefits, reducing land and water use, and waste...

In China, a handful of commercial projects already use Chinese DLE innovator Sunresin's technology.
More than 12 startups are pursuing new DLE processes, according to the article, "with the intent of commercial production as early as 2025."

And America's Department of Energy is also investing millions of dollars in new DLE tech "to extract lithium from geothermal brines in the U.S., such as the Salton Sea in California, which the National Renewable Energy Laboratory estimates could provide over 24,000 metric tons of lithium a year."

A year ago the Washington Post reported "there are only seven working offshore wind turbines in the entire United States," adding that a massive wind project south of Martha's Vineyard, Massachusetts "is years behind schedule amid regulatory delays and litigation from opponents."

But this week a local public radio station reported that electricity from America's first large-scale offshore wind project "is officially flowing into Massachusetts and helping to power the New England grid." The Vineyard Wind project achieved "first power" late Tuesday when one operating turbine near Martha's Vineyard delivered approximately five megawatts of electricity to the grid. The company said it expects to have five turbines operating at full capacity in early 2024... Once it's finished sometime in 2024, it will consist of 62 turbines spaced about a mile apart and rising more than 800 feet out of the water. The project will generate up to 800 megawatts of power, or about enough electricity for 400,000 homes in Massachusetts.

Another smaller project near Long Island, South Fork Wind, also began producing electricity in early December. When that project is complete, its 12 turbines will generate about 132 megawatts of power...

Massachusetts, in partnership with Rhode Island and Connecticut, is currently seeking bids for another 3,600 megawatts of offshore wind power... "This is a historic moment for the American offshore wind industry," wrote Gov. Maura Healey. "This is clean, affordable energy made possible by the many advocates, public servants, union workers, and business leaders who worked for decades to accomplish this achievement.
Last year America's seven offshore wind turbines generated "a paltry 42 megawatts," according to the article, "far less than the average natural gas power plant."

The CEO of one of the company's behind the project hailed the last 12 months as "a historic year defined by steel in the water and people at work."

Huawei's newest laptop runs on a chip made by Taiwan Semiconductor Manufacturing Co., a teardown of the device showed, quashing talk of another Chinese technological breakthrough. From a report: The Qingyun L540 notebook contains a 5-nanometer chip made by the Taiwanese company in 2020, around the time US sanctions cut off Huawei's access to the chipmaker, research firm TechInsights found after dismantling the device for Bloomberg News. That counters speculation that Huawei's mainland Chinese chipmaking partner, Semiconductor Manufacturing International Corp., may have achieved a major leap in fabrication technique.

Huawei caused a stir in the US and China last August when it released a smartphone with a 7nm processor made by Shanghai-based SMIC. A teardown by the Canada-based research outfit for Bloomberg News showed the Mate 60 Pro's chip was only a few years behind the cutting edge, a feat that US trade curbs were meant to prevent. That revelation spurred celebration across the Chinese tech scene, and a debate in the US about the effectiveness of sanctions.

An anonymous reader shares a report: The DeepMind robotics team has revealed three new advances that it says will help robots make faster, better, and safer decisions in the wild. One includes a system for gathering training data with a "Robot Constitution" to make sure your robot office assistant can fetch you more printer paper -- but without mowing down a human co-worker who happens to be in the way.

Google's data gathering system, AutoRT, can use a visual language model (VLM) and large language model (LLM) working hand in hand to understand its environment, adapt to unfamiliar settings, and decide on appropriate tasks. The Robot Constitution, which is inspired by Isaac Asimov's "Three Laws of Robotics," is described as a set of "safety-focused prompts" instructing the LLM to avoid choosing tasks that involve humans, animals, sharp objects, and even electrical appliances.

For additional safety, DeepMind programmed the robots to stop automatically if the force on its joints goes past a certain threshold and included a physical kill switch human operators can use to deactivate them. Over a period of seven months, Google deployed a fleet of 53 AutoRT robots into four different office buildings and conducted over 77,000 trials. Some robots were controlled remotely by human operators, while others operated either based on a script or completely autonomously using Google's Robotic Transformer (RT-2) AI learning model.