The gap between basic and advanced AI implementation in educational operations represents both a challenge and an opportunity.

Recap: Early last month, someone used the PDF format's JavaScript support to run Tetris inside what should normally be a static text document. Predictably, within days, a high school student upgraded the hack to run Doom within a PDF file. The same developer has now enhanced the code to run the entire Linux operating system. Barely a month after unveiling a port of Doom running inside a PDF, high school student and programmer "Ading2210" has successfully emulated Linux within the popular file format. Although performance is limited, the project redefines what's possible with PDF JavaScript tools. Users can try it here using Chromium browsers like Chrome, Edge, and Opera. The source code is available on the developer's GitHub page.LinuxPDF runs in a RISC-V emulator based on TinyEMU. Its inner workings closely resemble those of Ading2210's DoomPDF. For example, the inputs repeat the trick pioneered by the earlier Tetris PDF hack, reusing the Doom port's code. Users can click on virtual keys below the main screen, but most will likely prefer direct keyboard controls, which work by interpreting inputs in a text field.Although the PDF format was primarily designed to display text and images, it can also run JavaScript code. Adobe Acrobat includes the entire JavaScript specification, enabling features like 3D rendering, monitor detection, and HTTP requests.PDFs running in browsers use a more limited version, but it's good enough to run games and operating systems. Ading2210 discovered that an old version of Emscripten that targets asm.js instead of WebAssembly can compile C code to run within the file format.Like DoomPDF, the Linux emulation suffers from slow performance. Booting the kernel takes up to a full minute about 100 times longer than a traditional Linux system. According to Ading2210, this cannot, unfortunately, be fixed because Chromium uses a version of V8 that doesn't support the JIT compiler.The file system is 32-bit by default. However, users can build a 64-bit version from the source code by cloning the repository within a real Linux system, editing the "BITS" line, and downloading Emscripten version 1.39.20. Sadly, running the 64-bit version doubles the performance deficit.Users interested in a more practical Linux application for low-end hardware can try Ading2210's ChromeOS RMA Shim Bootloader. The script collection allows a full Debian distro to run on a Chromebook without modifying the firmware. The project also supports enrolled enterprise devices. // Related Stories

Elon Musk was one of the founding members of OpenAI, but made a sour exit before ChatGPT became a thing. The billionaire claims he wasnt happy with the non-profits pivot to a profit-chasing business model. A few days ago, Musk submitted a bid to buy OpenAIs non-profit arm for $97.4 billion, but now says he will pull the offer if the AI giant abandons its for-profit ambitions.If (the) OpenAI board is prepared to preserve the charitys mission and stipulate to take the for sale sign off its assets by halting its conversion, Musk will withdraw the bid, says a court filing submitted by the billionaires lawyer, as per Reuters.Recommended VideosMusks bid was surprising, though not entirely unexpected. Musk is currently entangled in a legal tussle with OpenAI, challenging the companys shift to a for-profit structure. OpenAI is aiming to spin its AI stack into a business entity, while the non-profit arm will retain a small stake in the business division. Swindler Elon Musk (@elonmusk) February 10, 2025When news of the acquisition offer broke, OpenAI chief, Sam Altman, responded that they would instead like to buy X at one-tenth the value of Musks bid. In subsequent press interactions, Altman made it clear that the company was not up for sale, adding that it was just another tactic to mess with the company.Interestingly, Reuters reported a few days ago that OpenAI hasnt formally received Musks $97.4 billion bid to acquire its non-profit arm. Altman labeled the attempt as ridiculous and reportedly told OpenAI employees that the company would reject the offer.But it seems this wasnt Musks first attempt at a takeover. In a couple of lengthy official statements published recently, OpenAI claimed Musk wanted to eventually merge the AI upstart with Tesla and foresaw a for-profit model, too. Altman also told Axios two days ago that Musk has been vying to take control of OpenAI for a long time.YUCK Sam Altman (@sama) February 11, 2025OpenAI is not for sale. OpenAIs mission is not for sale to say nothing of the fact that, like, a competitor who is not able to beat us in the market and you know, instead is just trying to say, like, Im gonna buy this with total disregard for the mission is a likely path there, Altman told the outlet.Musk owns his own AI company called xAI and already has a ChatGPT rival called Grok out in the market. Speaking virtually at the World Governments Summit in Dubai, Musk revealed that the latest version of xAIs chatbot, Grok-3, is in the late stages of development with improved reasoning capabilities.OpenAI has already teased its next-gen GPT-5 plans and says it will simplify the AI model strategy in the near future. The company is also working on in-house AI chips and the ambitious Stargate project worth $500 billion to boost AI development and help maintain US supremacy in the AI race.Editors Recommendations

One of the four Xs in 4X games like Civilization 7 is expand. You will never grow fast enough to compete with the other Leaders if you stick to just your starting capital city. An obvious but essential tip for Civilization 7 is to quickly expand out and found new towns to get more resources. As useful as towns are, you will eventually want to convert some of them into full cities for even greater benefits. The tutorials will tell you this is possible, but leaves out the important detail of exactly how. It won't take long, but will cost you plenty of gold.Recommended VideosDifficultyModerateDuration40 minutes2K GamesOnce you have sent a settler out and established a town, it will begin funneling some resources back to your capital. Until you upgrade them to towns, the only way to expand them is to purchase new infrastructure with gold. Towns don't require much attention, but you get more benefits such as creating military units and being able to accept more resource bonuses if you're willing to manage multiple cities. The only requirement you need when ready to upgrade towns to cities is decent chunk of gold.Step 1: Create a settler and have them establish a town on an eligible tile on the map.Step 2: Once the town is established and you want to convert it into a full city, click on the main town tile where you can pull up the list of purchase options.2K GamesRelatedStep 3: Tucked away at the very bottom border of the menu is the option to upgrade the town and how much gold it will cost.Be aware that the price to convert a town into a city will go up with each additional city you have.Editors Recommendations

The unsolicited offer has complicated OpenAI CEO Sam Altmans plans for the ChatGPT makers future.

MANY HANDS, LIGHT WORK Financially motivated hackers are helping their espionage counterparts and vice versa Two players who mostly worked independently are increasingly collaborative. Dan Goodin Feb 13, 2025 6:00 am | 4 Credit: Lino Mirgeler/picture alliance via Getty Images Credit: Lino Mirgeler/picture alliance via Getty Images Story textSizeSmallStandardLargeWidth *StandardWideLinksStandardOrange* Subscribers only Learn moreTheres a growing collaboration between hacking groups engaging in espionage on behalf of nation-states and those seeking financial gains through ransomware and other forms of cybercrime, researchers noted this week.There has always been some level of overlap between these two groups, but it has become more pronounced in recent years. On Tuesday, the Google-owned Mandiant security firm said the uptick comes amid tighter purse strings and as a means for concealing nation-state-sponsored espionage by making it blend in with financially motivated cyberattacks.Opportunities aboundModern cybercriminals are likely to specialize in a particular area of cybercrime and partner with other entities with diverse specializations to conduct operations, Mandiant researchers explained. The specialization of cybercrime capabilities presents an opportunity for state-backed groups to simply show up as another customer for a group that normally sells to other criminals. Purchasing malware, credentials, or other key resources from illicit forums can be cheaper for state-backed groups than developing them in-house, while also providing some ability to blend in to financially motivated operations and attract less notice."The report noted an increase in the sharing of malware between cybercrime groups and the governments of Russia, China, and Iran. The sharing goes both ways. Examples of threat actors working for these governments include:The use by the Russian-state hacking group APT44 of multiple pieces of crimeware with names including DarkCrystalRat, WarZone, and RadThief. APT44 has also used the hosting infrastructure of a well-known bulletproof host. Bulletproof hosts provide services that are tailored to cybercriminals to ensure their operations arent taken down by law enforcement.The use of the same RadThief malware by an Iranian-state actor tracked as UNC5203.A China-state espionage operator tracked as UNC2286 repeatedly using the SteamTrain ransomware along with a ransomware note from a different group tracked as DarkSide.On Thursday, researchers with the Symantec security firm reported on a collaboration that worked the other wayuse by the RA World ransomware group of a distinct toolset that previously has been seen used only in espionage operations by a China-linked threat group.The toolset, first spotted in July, was a variant of PlugX, a custom backdoor. Timestamps in the toolset were identical to those found by security firm Palo Alto Network in the Thor PlugX variant, which company researchers linked to a Chinese espionage group tracked under the names Fireant, Mustang Panda, and Earth Preta. The variant also had similarities to the PlugX type 2 variant found by security firm Trend Micro.Further espionage attacks involving the same PlugX variant occurred in August, when the attacker compromised the government of a southeastern European country. That same month, the attacker compromised a government ministry in a Southeast Asian country. In September 2024, the attacker compromised a telecoms operator in that region, and in January, the attacker targeted a government ministry in another Southeast Asian country.Symantec researchers have competing theories about the reason for this collaboration:There is evidence to suggest that this attacker may have been involved in ransomware for some time. In a report on RA World attacks, Palo Alto said that it had found some links to Bronze Starlight (aka Emperor Dragonfly), a China-based actor that deploys different ransomware payloads. One of the tools used in this ransomware attack was a proxy tool called NPS, which was created by a China-based developer. This has previously been used by Bronze Starlight. SentinelOne, meanwhile, reported that Bronze Starlight had been involved in attacks involving the LockFile, AtomSilo, NightSky, and LockBit ransomware families.It is unclear why an actor who appears to be linked to espionage operations is also mounting a ransomware attack. While this is not unusual for North Korean threat actors to engage in financially motivated attacks to subsidize their operations, there is no similar history for China-based espionage threat actors, and there is no obvious reason why they would pursue this strategy.Another possibility is that the ransomware was used to cover up evidence of the intrusion or act as a decoy to draw attention away from the true nature of the espionage attacks. However, the ransomware deployment was not very effective at covering up the tools used in the intrusion, particularly those linking it back to prior espionage attacks. Secondly, the ransomware target was not a strategically significant organization and was something of an outlier compared to the espionage targets. It seems unusual that the attacker would go to such lengths to cover up the nature of their campaign. Finally, the attacker seemed to be serious about collecting a ransom from the victim and appeared to have spent time corresponding with them. This usually wouldnt be the case if the ransomware attack was simply a diversion.The most likely scenario is that an actor, possibly one individual, was attempting to make some money on the side using their employers toolkit.Tuesdays report from Mandiant also noted the use of state-sponsored malware by crime groups. Mandiant researchers also reported observing what they believe are Dual Motive groups that seek both financial gain and access for espionage.Dan GoodinSenior Security EditorDan GoodinSenior Security Editor Dan Goodin is Senior Security Editor at Ars Technica, where he oversees coverage of malware, computer espionage, botnets, hardware hacking, encryption, and passwords. In his spare time, he enjoys gardening, cooking, and following the independent music scene. Dan is based in San Francisco. Follow him at here on Mastodon and here on Bluesky. Contact him on Signal at DanArs.82. 4 Comments

Over the past decade, corn and soybeans arent the only things sprouting from fertile Iowa soil. Data centers have been popping up with growing regularity. Tech giants like Apple, Google, Meta, and Microsoft have flocked to the Hawkeye State due to ample land, low energy costs, minimal earthquake risk, and generous tax incentives.But as climate change accelerates and water tables drop in some regions, a critical debate is surfacing. Groundwater in Iowa is not evenly distributed. Data centers in one place can be very different than in another, observes Keith Schilling, state geologist and director of the Iowa Geological Survey at the University of Iowa.Not surprisingly, the growing demand for water is trickling into policy, consumption patterns, environmental impacts and costs. Amid competing demands -- agriculture, business and residential use -- Iowa officials are taking notice. With increased demand and continued drought, aquifers arent being recharged as they were in the past, Schilling notes.Into the FlowAt present, 34 data centers exist across Iowa. Each day, these facilities consume somewhere between 300,000 gallons and 1.25 million gallons of water for cooling. At any given location, they typically account for about 2% to 8% of total water consumption -- though some facilities have agreements to use more water, if necessary.Related:Much of this water comes from underground aquifers. Under normal circumstances, theres sufficient rainfall to support agriculture, manufacturing and home use, but for more than a decade most of the state has endured a drought. This has translated into a need to pump additional water for crops, food processing, and ethanol production. According to Iowa Environmental Council, some wells are now operating at 20% of their original capacity.To be sure, the water situation varies considerably across the state -- and even within regions. For example, the state is bounded by the Missouri River and Mississippi River and other waterways run through Iowa. The eastern third of the state has abundant shallow carbonate aquifers that are recharged every year with precipitation, Schilling says. In central and western Iowa, the conditions are less favorable. The groundwater systems are shallower and more vulnerable.Twenty of the states data centers are in the Des Moines area, which is near the center of Iowa. So far, this part of the state has avoided problems associated with droughts and water shortages. But as newer facilities come online, water consumption increases and drought lingers, questions about water availability are growing -- particularly in more vulnerable areas in the western part of the state.Related:In many cases, companies looking to build data centers are using criteria other than water availability to select a site, Schilling says. They are often more interested in the surrounding infrastructure.Beneath the SurfaceBalancing water use across agriculture, manufacturing firms and tech companies -- while keeping rates down for residential users -- is a balancing act based in both economics and sustainability. When fixed costs can be spread amongst more users, everyone experiences lower rates -- including residential customers, says Roy Hesemann, utilities director for Cedar Rapids.Located toward the eastern portion of the state -- adjacent to the Cedar River -- the city has sufficient water and energy resources to support major data centers, Hesemann says. Cedar Rapids recently approved a $576 million Google data center that will use 200,000 to 1 million gallons of water daily and pull 25 megawatts of electricity. The project will generate $1 billion in local property taxes over 20 years (with about $529 million flowing back to Google).Related:The facility wont place additional pressure on supply or impact water standards. However, It will require us to accelerate timelines for planned expansion at the Northwest Water Treatment Plant to meet future capacity needs, Hesemann says. For now, the citys water rates rank in the middle for Iowa. Gaining additional efficiency is important. The city is exploring ways to reuse water, including discharge from data centers.Google is planning another $1 billion data center in Council Bluffs, which sits in the more arid western portion of the state. Once completed, the facility will add capacity to two other facilities built in 2012 and 2015. The expansion will result in a total of three datacenters that comprise nearly 3 million square feet of space split among 3 buildings. Google has invested over $5 billion in the region since 2007.Although these projects make economic sense, observers such as Schilling are taking notice and advocating for a better understanding of how data centers impact water consumption in Iowa. With ongoing drought conditions and growing demand for limited water resources, some aquifers are not being recharged adequately, he explains. Some areas may not be ideal locations for data centers.Data StreamsThe Iowa legislature is taking notice. In 2024, it designated $250,000 to map aquifers and study groundwater levels to gain a better understanding of how various user groups -- including data centers -- impact recharge rates and water levels. Governor Kim Reynolds described the funding as crucial for the development of models for budgeting this state's water resources."A growing dependence on data centers may be cause for concern, but digital technologies might also provide answers for Iowa. In July 2021, Iowa State University announced that it had landed $20 million grant from the US federal government to establish an AI Institute for Resilient Agriculture. It is developing digital twins, robotics, drones and connected field sensors that reduce water and chemical use while boosting crop yields.It makes sense for companies to establish data centers and other facilities in Iowa because of land availability, water resources, financial incentives and renewable energy, says Soumik Sarkar, a professor of mechanical engineering and computer science at Iowa State University. At the same time, we have witnessed a 20% to 30% increase in water demand for large AI companies and the state is coping with a drought. So, we must find ways to manage resources better.Another five-year, $7 million federal government funded project called COALESCE (COntext Aware LEarning for Sustainable CybEr-agricultural systems) is helping researchers at Iowa State University study ways to embed digital technologies, including AI, deeper into the food production system. AI and other tools can help us optimize processes, reduce pesticides and pollution and maximize our water systems, Sarkar states.Its important to understand how data centers and existing infrastructure impact water use -- and aquifers in Iowa, Schilling says. We dont want to reach a point where shortages occur, and people claim that the situation has taken them by surprise. We have the technology and tools to manage water resources effectively. We must use them to determine where to locate data centers and other facilities.

From COVID-19 to war in Ukraine, and more, the past five years have brought cybersecurity to mainstream attention.The US Department of Defense recently hosted an international exchange on shaping cybersecurity workforce, following the publication of its 2023 strategy to align the department's efforts to identify, recruit, develop, and retain a data-literate and technology-adept cyber workforce. These actions, among similar developments globally, provide insights into some of the challenges that CISOs and cybersecurity teams will face in the coming years.In practice, 2025 is likely to see growing importance of and demand for CISOs. The growing threat ofglobal and regional political instability, paired with the increasing capabilities of violent extremist organizations and crime groups seeking to cause harm, means that access to data will become a key component of global power for both state and non-state actors -- all of which will require greater vigilance from cyber teams.Another trend driving cyber threats is the technological arms race. Driven by advances in quantum computing and artificial intelligence, the race between cyber exploiters and victims has further intensified. Cybersecurity and AI are now bipartisan national security issues and crucial components of Americas competitive advantage. Simultaneously, increasing tools and incentives for cybercriminals and advanced persistent threats (APTs) will continue to raise the stakes for private sector firms. The rise of zero-day attacks only further highlights the evolving tactics of cyber adversaries, and CISOs must remain vigilant to protect their organizations.Related:This is set against a shift in current political landscape in the US, with the incoming administration potentially marking a significant change in the cybersecurity demands on firms as they seek to reduce red tape.Heres a look at the top cybersecurity trends that will shape 2025 and beyond.1. Navigating SEC cybersecurity disclosure rulesIn 2024, new SEC cybersecurity disclosure rules led to a significant increase in the public reporting of incidents. However, the often-vague nature of these disclosures and their limited detail on impact left investors seeking greater clarity.While the incoming administration may consider rescinding these requirements to reduce regulatory burdens, it is more likely that the current status quo will persist through 2025. CISOs should take a proactive approach by analyzing disclosures made in 2024 to understand how they were received and pre-plan the level of disclosure their organization is prepared to make. This will help mitigate risks and ensure transparency while complying with existing requirements.Related:2. Understanding AIs complex roleArtificial intelligence will remain a focal point for cybersecurity teams in 2025. AIs adversarial uses, as highlighted by the FBI at RSA in 2024, include creating undetectable malware, automating reconnaissance, and executing deepfake scams. Simultaneously, organizations are pursuing the AI dream to unlock significant business benefits, often without fully considering security implications.To ensure safe usage of AI technology, CISOs must engage at the planning stages of adoption to ensure security is integrated rather than treated as an afterthought. Boards now expect clear strategies to address AI-related risks, including sophisticated phishing and social engineering attacks enabled by AI.CISOs must balance fostering innovation with maintaining robust security measures. They can do this by investing heavily in protecting their digital systems, physical assets and workforce from adversaries. By implementing software solutions capable of detecting cyber threats, restricting access to buildings, and safeguarding sensitive employee information -- CISOs can take the necessary steps to fortify their defenses.Related:3. Strengthening security culture to mitigate human errorDespite technological advancements, human actions -- whether through unintentional errors or deliberate breaches -- remain a primary cause of security incidents. In fact, up to 95% of successful security attacks result from human error.As technical solutions alone are insufficient to protect organizations, fostering a robust security culture becomes essential. Embedding security awareness and proactive behaviors into the organizational culture ensures that every employee understands their role in safeguarding sensitive information and digital assets. This human-centric approach provides a vital first line of defense, empowering individuals to act as security champions and take a proactive role in mitigating associated risks.4. Adapting to AI regulationsState-level AI regulations in the US will present significant challenges for CISOs in 2025. States such as Colorado, California, and Utah have already passed private-sector AI rules with varying effective dates, creating a complex compliance landscape. The absence of a pre-emptive federal approach means that organizations must navigate a patchwork of reporting, assessment, and governance requirements.Fortunately, frameworks like NISTs AI RMF and ISO 42001 offer a common foundation for compliance, enabling organizations to demonstrate their commitment to ethical and secure AI practices. Preparing for these requirements, along with global mandates such as the EU AI Act, will be a critical focus for cybersecurity teams in the coming year.5. Preparing for post-quantum cryptographyThe release of NISTs post-quantum encryption tools marks a pivotal moment for cybersecurity planning.The harvest now, decrypt later strategy employed by adversaries underscores the urgency of transitioning to post-quantum cryptography. Organizations must define multiyear strategies to implement these new standards to safeguard sensitive data against future quantum threats. Early adopters of post-quantum cryptography demonstrate not only technical readiness but also a commitment to customer data protection. CISOs who act decisively in 2025 will position their organizations as leaders in cybersecurity resilience.As we look ahead to 2025, the challenges facing CISOs, and cybersecurity teams are complex and multifaceted. From navigating SEC disclosure requirements and managing AI-related risks to strengthening security culture and preparing for post-quantum threats, proactive planning and strategic action are essential.By staying ahead of these trends, organizations can strengthen their defenses, protect critical assets, and maintain trust in an increasingly interconnected and digital era.

This article is from The Spark, MIT Technology Reviews weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here. A few weeks ago, a fire broke out at the Moss Landing Power Plant in California, the worlds largest collection of batteries on the grid. Although the flames were extinguished in a few days, the metaphorical smoke is still clearing. Some residents in the area have reported health issues that they claim are related to the fire, and some environmental tests revealed pollutants in the water and ground near where the fire burned. One group has filed a lawsuit against the company that owns the site. In the wake of high-profile fires like Moss Landing, there are very understandable concerns about battery safety. At the same time, as more wind, solar power, and other variable electricity sources come online, large energy storage installations will be even more crucial for the grid. Lets catch up on what happened in this fire, what the lingering concerns are, and what comes next for the energy storage industry. The Moss Landing fire was spotted in the afternoon on January 16, according to local news reports. It started small but quickly spread to a huge chunk of batteries at the plant. Over 1,000 residents were evacuated, nearby roads were closed, and a wider emergency alert warned those nearby to stay indoors. The fire hit the oldest group of batteries installed at Moss Landing, a 300-megawatt array that came online in 2020. Additional installations bring the total capacity at the site to about 750 megawatts, meaning it can deliver as much energy to the grid as a standard coal-fired power plant for a few hours at a time. According to a statement that site owner Vistra Energy gave to the New York Times, most of the batteries inside the affected building (the one that houses the 300MW array) burned. However, the company doesnt have an exact tally, because crews are still prohibited from going inside to do a visual inspection. This isnt the first time that batteries at Moss Landing have caught firethere have been several incidents at the plant since it opened. However, this event was much more significant than previous fires, says Dustin Mulvaney, a professor of environmental studies at San Jose State University, whos studied the plant. Residents are worried about the potential consequences.The US Environmental Protection Agency monitored the nearby air for hydrogen fluoride, a dangerous gas that can be produced in lithium-ion battery fires, and didnt detect levels higher than Californias standards. But some early tests detected elevated levels of metals including cobalt, nickel, copper, and manganese in soil around the plant. Tests also detected metals in local drinking water, though at levels considered to be safe. Citing some of those tests, a group of residents filed a lawsuit against Vistra last week, alleging that the company (along with a few other named defendants) failed to implement adequate safety measures despite previous incidents at the facility. The suits legal team includes Erin Brockovich, the activist famous for her work on a 1990s case against Pacific Gas & Electric Company involving contaminated groundwater from oil and gas equipment in California. The lawsuit, and Brockovichs involvement in particular, raises a point that I think is worth recognizing here: Technologies that help us address climate change still have the potential to cause harm, and taking that seriously is crucial. The oil and gas industry has a long history of damaging local environments and putting people in harms way. Thats evident in local accidents and long-term pollution, and in the sense that burning fossil fuels drives climate change, which has widespread effects around the world. Low-carbon energy sources like wind, solar, and batteries dont add to the global problem of climate change. But many of these projects are industrial sites, and their effects can still be felt by local communities, especially when things go wrong as they did in the Moss Landing fire. The question now is whether those concerns and lawsuits will affect the industry more broadly. In a news conference, one local official called the fire a Three Mile Island event for this industry, referring to the infamous 1979 accident at a Pennsylvania nuclear power plant. That was a turning point for nuclear power, after which public support declined sharply. With the growing number of electric vehicles and batteries for energy storage on the grid, more high-profile fires have hit the news, like last years truck fire in LA, the spate of e-bike battery fires in New York City, or one at a French recycling plant last year. Battery energy storage systems are complex machines, Mulvaney says. Complex systems have a lot of potential failures. When it comes to large grid-scale installations, battery safety has already improved since Moss Landing was built in 2020, as Canary Medias Julian Spector points out in a recent story. One reason is that many newer sites use a different chemistry thats considered safer. Newer energy storage facilities also tend to isolate batteries better, so small fires wont spread as dramatically as they did in this case. Theres still a lot we dont know about this fire, particularly when it comes to how it started. Learning from the results of the ongoing investigations will be important, because we can only expect to see more batteries coming online in the years ahead. In 2023, there were roughly 54 gigawatts worth of utility-scale batteries on the grid globally. If countries follow through on stated plans for renewables, that number could increase tenfold by the end of the decade. Energy storage is a key tool in transforming our grid and meeting our climate goals, and the industry is moving quickly. Safety measures need to keep up. Now read the rest of The Spark Related reading E-bike battery fires, including ones started by delivery drivers vehicles, have plagued New York City. A battery-swapping system could help address the problem. Insulating materials layered inside EV batteries could help reduce fire risk. A company making them just got a big boost in the form of a loan from the US Department of Energy. New chemistries, like iron-air batteries, promise safer energy storage. Read our profile of Form Energy, which we named one of our 15 Climate Tech Companies to Watch in 2024. Keeping up with climate Data centers are expected to be a major source of growth in electricity demand. Being flexible may help utilities meet that demand, according to a new study. (Inside Climate News) The worlds first lab-grown meat for pets just went on sale in the UK. Meatly is selling limited quantities of its treats, which are a blend of plant-based ingredients and cultivated chicken cells. (The Verge) Kore Power scrapped plans for a $1.2 billion battery plant in Arizona, but the company isnt giving up just yet. The new CEO said the new plan is to look for an existing factory that can be transformed into a battery manufacturing facility. (Canary Media) The auto industry is facing a conundrum: Customers in the US want bigger vehicles, but massive EVs might not make much economic sense. New extended-range electric vehicles that combine batteries and a gas-powered engine that acts as a generator could be the answer. (Heatmap) Officials at the National Oceanic and Atmospheric Administration were told to search grants for words related to climate change. Its not clear what comes next. (Axios) It might be officially time to call it on the 1.5 C target. Two new studies suggest that the world has already entered into the runway to surpass the point where global temperatures increase 1.5 C over preindustrial levels. (Bloomberg) States are confused over a Trump administration order to freeze funding for EV chargers. Some have halted work on projects under the $5 billion program, while others are forging on. (New York Times) Cold weather can affect the EV batteries. Criticisms likely portray something way worse than the reality, but in any case, heres how to make the most of your EV in the winter. (Canary Media)

Jules Rodriguez lost his voice in October of last year. His speech had been deteriorating since a diagnosis of amyotrophic lateral sclerosis (ALS) in 2020, as the muscles in his head and neck progressively weakened along with those in the rest of his body. By 2024, doctors were worried that he might not be able to breathe on his own for much longer. So Rodriguez opted to have a small tube inserted into his windpipe to help him breathe. The tracheostomy would extend his life, but it also brought an end to his ability to speak. A tracheostomy is a scary endeavor for people living with ALS, because it signifies crossing a new stage in life, a stage that is close to the end, Rodriguez tells me using a communication device. Before the procedure I still had some independence, and I could still speak somewhat, but now I am permanently connected to a machine that breathes for me. Rodriguez and his wife, Maria Fernandez, who live in Miami, thought they would never hear his voice again. Then they re-created it using AI. After feeding old recordings of Rodriguezs voice into a tool trained on voices from film, television, radio, and podcasts, the couple were able to generate a voice clonea way for Jules to communicate in his old voice. Hearing my voice again, after I hadnt heard it for some time, lifted my spirits, says Rodriguez, who today communicates by typing sentences using a device that tracks his eye movements, which can then be spoken in the cloned voice. The clone has enhanced his ability to interact and connect with other people, he says. He has even used it to perform comedy sets on stage. Rodriguez is one of over a thousand people with speech difficulties who have used the voice cloning tool since ElevenLabs, the company that developed it, made it available to them for free. Like many new technologies, the AI voice clones arent perfect, and some people find them impractical in day-to-day life. But the voices represent a vast improvement on previous communication technologies and are already improving the lives of people with motor neuron diseases, says Richard Cave, a speech and language therapist at the Motor Neuron Disease Association in the UK. This is genuinely AI for good, he says. Cloning a voice Motor neuron diseases are a group of disorders in which the neurons that control muscles and movement are progressively destroyed. They can be difficult to diagnose, but typically, people with these disorders start to lose the ability to move various muscles. Eventually, they can struggle to breathe, too. There is no cure. Rodriguez started showing symptoms of ALS in the summer of 2019. He started losing some strength in his left shoulder, says Fernandez, who sat next to him during our video call. We thought it was just an old sports injury. His arm started to get thinner, too. In November, his right thumb stopped working while he was playing video games. It wasnt until February 2020, when Rodriguez saw a hand specialist, that he was told he might have ALS. He was 35 years old. It was really, really, shocking to hear from somebody you see about your hand, says Fernandez. That was a really big blow. Like others with ALS, Rodriguez was advised to bank his voiceto tape recordings of himself saying hundreds of phrases. These recordings can be used to create a banked voice to use in communication devices. The result was jerky and robotic. Its a common experience, says Cave, who has helped 50 people with motor neuron diseases bank their voices. When I first started at the MND Association [around seven years ago], people had to read out 1,500 phrases, he says. It was an arduous task that would take months. And there was no way to predict how lifelike the resulting voice would beoften it ended up sounding quite artificial. It might sound a bit like them, but it certainly couldnt be confused for them, he says. Since then, the technology has improved, and for the last year or two the people Cave has worked with have only needed to spend around half an hour recording their voices. But though the process was quicker, he says, the resulting synthetic voice was no more lifelike. Then came the voice clones. ElevenLabs has been developing AI-generated voices for use in films, televisions, and podcasts since it was founded three years ago, says Sophia Noel, who oversees partnerships between the company and nonprofits. The companys original goal was to improve dubbing, making voice-overs in a new language seem more natural and less obvious. But then the technical lead of Bridging Voice, an organization that works to help people with ALS communicate, told ElevenLabs that its voice clones were useful to that group, says Noel. Last August, ElevenLabs launched a program to make the technology freely available to people with speech difficulties. Suddenly, it became much faster and easier to create a voice clone, says Cave. Instead of having to record phrases, users can instead upload voice recordings from past WhatsApp voice messages or wedding videos, for example. You need a minimum of a minute to make anything, but ideally you want around 30 minutes, says Noel. You upload it into ElevenLabs. It takes about a week, and then it comes out with this voice. Rodriguez played me a statement using both his banked voice and his voice clone. The difference was stark: The banked voice was distinctly unnatural, but the voice clone sounded like a person. It wasnt entirely naturalthe words came a little fast, and the emotive quality was slightly lacking. But it was a huge improvement. The difference between the two is, as Fernandez puts it, like night and day. The ums and ers Cave started introducing the technology to people with MND a few months ago. Since then, 130 of them have started using it, and the feedback has been unremittingly good, he says. The voice clones sound far more lifelike than the results of voice banking. They [include] pauses for breath, the ums, the ers, and sometimes there are stammers, says Cave, who himself has a subtle stammer. That feels very real to me, because actually I would rather have a synthetic voice representing me that stammered, because thats just who I am. Joyce Esser is one of the 130 people Cave has introduced to voice cloning. Esser, who is 65 years old and lives in Southend-on-Sea in the UK, was diagnosed with bulbar MND in May last year. Bulbar MND is a form of the disease that first affects muscles in the face, throat, and mouth, which can make speaking and swallowing difficult. Esser can still talk, but slowly and with difficulty. Shes a chatty person, but she says her speech has deteriorated quite quickly since January. We communicated via a combination of email, video call, speaking, a writing board, and text-to-speech tools. To say this diagnosis has been devastating is an understatement, she tells me. Losing my voice has been a massive deal for me, because its such a big part of who I am. Joyce Esser and her husband Paul on holiday in the Maldives.COURTESY OF JOYCE ESSER Esser has lots of friends all over the country, Paul Esser, her husband of 38 years, tells me. But when they get together, they have a rule: Dont talk about it, he says. Talking about her MND can leave Joyce sobbing uncontrollably. She had prepared a box of tissues for our conversation. Voice banking wasnt an option for Esser. By the time her MND was diagnosed, she was already losing her ability to speak. Then Cave introduced her to the ElevenLabs offering. Esser had a four-and-a-half-minute-long recording of her voice from a recent local radio interview and sent it to Cave to create her voice clone. When he played me my AI voice, I just burst into tears, she says. ID GOT MY VOICE BACK!!!! Yippeeeee! We were just beside ourselves, adds Paul. We thought wed lost [her voice] forever. Hearing a lost voice can be an incredibly emotional experience for everyone involved. It was bittersweet, says Fernandez, recalling the first time she heard Rodriguezs voice clone. At the time, I felt sorrow, because [hearing the voice clone] reminds you of who he was and what weve lost, she says. But overwhelmingly, I was just so thrilled it was so miraculous. Rodriguez says he uses the voice clone as much as he can. I feel people understand me better compared to my banked voice, he says. People are wowed when they first hear it as I speak to friends and family, I do get a sense of normalcy compared to when I just had my banked voice. Cave has heard similar sentiments from other people with motor neuron disease. Some [of the people with MND Ive been working with] have told me that once they started using ElevenLabs voices people started to talk to them more, and that people would pop by more and feel more comfortable talking to them, he says. Thats important, he stresses. Social isolation is common for people with MND, especially for those with advanced cases, he says, and anything that can make social interactions easier stands to improve the well-being of people with these disorders: This is something that [could] help make lives better in what is the hardest time for them. I dont think I would speak or interact with others as much as I do without it, says Rodriguez. A very slow game of Ping-Pong But the tool is not a perfect speech aid. In order to create text for the voice clone, words must be typed out. There are lots of devices that help people with MND to type using their fingers or eye or tongue movements, for example. The setup works fine for prepared sentences, and Rodriguez has used his voice clone to deliver a comedy routinesomething he had started to do before his ALS diagnosis. As time passed and I began to lose my voice and my ability to walk, I thought that was it, he says. But when I heard my voice for the first time, I knew this tool could be used to tell jokes again. Being on stage was awesome and invigorating, he adds. Jules Rodriguez performs his comedy set on stage.DAN MONO FROM DART VISION But typing isnt instant, and any conversations will include silent pauses. Our arguments are very slow paced, says Fernandez. Conversations are like a very slow game of Ping-Pong, she says. Joyce Esser loves being able to re-create her old voice. But she finds the technology impractical. Its good for pre-prepared statements, but not for conversation, she says. She has her voice clone loaded onto a phone app designed for people with little or no speech, which works with ElevenLabs. But it doesnt allow her to use swipe typinga form of typing she finds to be quicker and easier. And the app requires her to type sections of text and then upload them one at a time, she says, adding: Id just like a simple device with my voice installed onto it that I can swipe type into and have my words spoken instantly. For the time being, her first choice communication device is a simple writing board. Its quick and the listener can engage by reading as I write, so its as instant and inclusive as can be, she says. Esser also finds that when she uses the voice clone, the volume is too low for people to hear, and it speaks too quickly and isnt expressive enough. She says shed like to be able to use emojis to signal when shes excited or angry, for example. Rodriguez would like that option too. The voice clone can sound a bit emotionally flat, and it can be difficult to convey various sentiments. The issue I have is that when you write something long, the AI voice almost seems to get tired, he says. We appear to have the authenticity of voice, says Cave. What we need now is the authenticity of delivery. Other groups are working on that part of the equation. The Scott-Morgan Foundation, a charity with the goal of making new technologies available to improve the well-being of people with disorders like MND, is working with technology companies to develop custom-made systems for 10 individuals, says executive director LaVonne Roberts. The charity is investigating pairing ElevenLabs voice clones with an additional technology hyperrealistic avatars for people with motor neuron disease. These twins look and sound like a person and can speak from a screen. Several companies are working on AI-generated avatars. The Scott-Morgan Foundation is working with D-ID. Creating the avatar isnt an easy process. To create hers, Erin Taylor, who was diagnosed with ALS when she was 23, had to speak 500 sentences into a camera and stand for five hours, says Roberts. We were worried it was going to be impossible, she says. The result is impressive. Her mom told me, Youre starting to capture [Erins] smile, says Roberts. That really hit me deeper and heavier than anything. Taylor showcased her avatar at a technology conference in January with a pre-typed speech. Its not clear how avatars like these might be useful on a day-to-day basis, says Cave: The technology is so new that were still trying to come up with use cases that work for people with MND. The question is how do we want to be represented? Cave says he has seen people advocate for a system where hyperrealistic avatars of a person with MND are displayed on a screen in front of the persons real face. I would question that right from the start, he says. Both Rodriguez and Esser can see how avatars might help people with MND communicate. Facial expressions are a massive part of communication, so the idea of an avatar sounds like a good idea, says Esser. But not one that covers the users face you still need to be able to look into their eyes and their souls. The Scott-Morgan Foundation will continue to work with technology companies to develop more communication tools for people who need them, says Roberts. And ElevenLabs plans to partner with other organizations that work with people with speech difficulties so that more of them can access the technology. Our goal is to give the power of voice to 1 million people, says Noel. It really does change the game for us, says Fernandez. It doesnt take away most of the things we are dealing with, but it really enhances the connection we can have together as a family.