#BrainUp Daily Tech News – (Friday, April 10ᵗʰ)
Welcome to today’s curated collection of interesting links and insights for 2026/04/10. Our Hand-picked, AI-optimized system has processed and summarized 34 articles from all over the internet to bring you the latest technology news.
As previously aired🔴LIVE on Clubhouse, Chatter Social, Instagram, Twitch, X, YouTube, and TikTok.
Also available as a #Podcast on Apple 📻, Spotify🛜, Anghami, and Amazon🎧 or anywhere else you listen to podcasts.
1. Japan to ban gene-edited embryos aimed at creating “designer babies”
Japan is set to ban research and treatments that genetically modify human fertilized eggs using #genome editing and then implant those embryos into human or animal uteruses to produce a child. A bill to outlaw these activities was approved by the Cabinet on Friday. The measure targets attempts to create so called “designer babies” by preventing the use of gene edited embryos for birth. It links regulation of #genome editing to a clear restriction on implantation aimed at childbirth, establishing a legal prohibition rather than allowing such work under research or clinical pathways.
2. Watch Artemis 2 race back to Earth in this telescope livestream tonight
A Virtual Telescope Project livestream will attempt to track @NASA’s Artemis 2 #Orion spacecraft as it speeds back toward Earth, offering a possible last deep space view before reentry, weather permitting. Hosted by astrophysicist @Gianluca Masi, the broadcast is scheduled to begin at 10:45 p.m. EDT on April 9 (0245 GMT on April 10) and will stream on Space.com and VTP’s WebTV. Astronomers using a network of robotic telescopes in Italy will try to follow Orion as a faint, fast moving point of light against background stars. Artemis 2 launched April 1 with four astronauts for a historic crewed flight around the moon, the first beyond low Earth orbit since Apollo 17 in 1972, and is now nearing its final return phase. Splashdown is expected Friday evening (April 10) in the Pacific Ocean off the coast of San Diego, with additional updates available via Space.com’s Artemis 2 live blog.
3. OpenAI: Hey, We Also Have a New Tool That Is So Scarily Powerful We Can’t Release It
According to Axios, @OpenAI is finishing a new #cybersecurity service it says is too powerful for broad release and will instead be offered as a separate product to a limited set of partners to help bolster defenses. The report says it is not a new model and is unrelated to the company’s planned upcoming release, Spud, with few other details disclosed. The piece argues this move looks like a response to rival @Anthropic’s rollout of its Mythos model for finding and fixing #security vulnerabilities, and suggests OpenAI is trying to keep pace in the hype cycle, noting it already runs an invitation-only “Trusted Access for Cyber” pilot offering more permissive models after GPT-5.3-Codex. It also questions attention-grabbing claims in the space, citing that Anthropic’s boasts about long-missed flaws may be exaggerated and framing OpenAI’s positioning as marketing catch-up rather than demonstrated breakthroughs. Separately, it points to OpenAI’s own ambitious projections, reporting it forecast $102 billion in advertising sales by 2030 versus a projected $2.5 billion this year, calling the gap another example of aggressive hype.
4. AI poses an increasing threat to entry-level jobs, experts warn
AI technologies are increasingly automating entry-level jobs, raising concerns among experts about future employment prospects for new workforce entrants. Research shows that AI tools can perform many routine tasks traditionally done by individuals starting their careers, particularly in sectors like customer service, data entry, and administrative roles. This shift could reduce the availability of opportunities for young people and those without advanced skills to gain workplace experience and develop essential competencies. Experts argue that the growing integration of #AI in workplaces demands new training programs and policies to support workforce adaptation and prevent deepening inequalities. Addressing these challenges is crucial to ensure that AI complements rather than replaces human roles, maintaining inclusive job markets.
5. FBI Extracts Suspect’s Deleted Signal Messages Saved in iPhone Notification Database
The @FBI forensically extracted copies of incoming #Signal messages from a defendant’s iPhone even after the app had been deleted, because the message content was stored in the device’s push notification database, according to multiple people who attended FBI testimony in a recent trial. The case concerned a group accused of setting off fireworks and vandalizing property at the ICE Prairieland Detention Facility in Alvarado, Texas in July, with one person alleged to have shot a police officer in the neck, and it was described as the first time authorities charged people for alleged “Antifa” activities after @Donald Trump designated the umbrella term a terrorist organization. The episode illustrates how #forensic extraction with physical access and specialized software can recover sensitive data derived from secure messaging apps from unexpected system locations rather than the app itself. It also underscores the importance of Signal’s setting that prevents message content from appearing in push notifications for users who need to reduce such exposure.
6. Turkey to race ahead of EU on battery storage amid fossil fuel crisis
Turkey has approved more #battery storage for its grid since 2022 than any EU member state, signalling a rapid clean-technology buildout as fossil fuel instability persists. The climate thinktank Ember reports more than 33GW approved in Turkey since 2022, versus about 12-13GW total planned and operational in early European leaders such as Germany and Italy, with Turkey’s surge driven by a 2022 rule granting preferential grid access to renewables paired with equal storage. Of 221GW of submitted storage applications, 33GW has been approved, mostly one-hour systems totalling 37GWh, and an Ember analyst said this policy created a “massive investment signal” that could underpin a regional clean energy hub if delivered. Analysts link the boom to steep cost declines, with @Greg Nemet noting solar and batteries have fallen by nearly 90% in price over the past decade, while Europe’s push for grids and storage has intensified after the Iran war triggered a new fossil fuel crisis. Turkey already gets about a fifth of its power from wind and solar but still heavily supports coal, and while it targets 120GW of wind and solar by 2035, last year’s additions lagged what is needed to stay on track as the country prepares to host #Cop31 in Antalya.
7. This Diet–Gut Interaction Could Transform Fat Into a Calorie-Burning Machine
Scientists have discovered a powerful new link between diet, gut bacteria, and fat metabolism, showing that certain foods can trigger microbes in the gut to reprogram fat cells from storing energy to burning it, at least in mice. The key concept here is the difference between #WhiteFat and #BeigeFat: white fat stores calories, while beige fat burns them to produce heat and boost metabolism. The breakthrough reveals that a low-protein diet activates specific gut microbes, which then release chemical signals that travel through the body and flip this metabolic switch, effectively turning fat into a “calorie-burning machine.”
The mechanism is especially important: it’s not just diet alone, but a #GutMicrobiome interaction, meaning your gut bacteria act like interpreters of what you eat. These microbes trigger a two-step biological process involving bile acids and a hormone called #FGF21, which together instruct fat cells to start burning energy instead of storing it. This introduces a deeper concept in biology called metabolic plasticity, where fat tissue is not fixed but can change function depending on internal signals. In experiments, mice with the right bacteria gained less weight, had better blood sugar control, and improved cholesterol levels, while mice without gut microbes showed none of these effects even on the same diet.
However, the study comes with a critical limitation: this is still early-stage research in mice, and the specific low-protein diet used is not recommended for humans, meaning this is not a practical diet hack yet. Instead, the real implication is pharmaceutical, where scientists aim to replicate these microbial signals as treatments for obesity and metabolic diseases rather than relying on extreme diets. Ultimately, the discovery reframes metabolism itself: your body is not just passively processing food, but actively negotiating with your gut microbes, turning diet into biochemical instructions that can reshape how energy is stored, burned, and regulated.
8. Bessent, Powell warn bank CEOs about Anthropic model risks: Bloomberg News reports
The article reports that financial leaders Neil Bessent and Jerome Powell have cautioned bank CEOs about the risks associated with Anthropic’s AI models. They emphasize that while #ArtificialIntelligence can offer innovations and efficiencies in banking, it also presents significant operational and reputational risks. The warnings include concerns about the models’ reliability, potential biases, and the challenge of regulating rapidly evolving AI technologies. This reflects a growing emphasis on cautious integration of AI tools in the financial sector, highlighting the need for robust risk assessment frameworks. The advisory aims to prompt banking leaders to implement safeguards when adopting Anthropic’s AI to protect institutions and customers.
9. Amazon CEO Andy Jassy’s pay rose to $2.1 million in 2025 as security and travel costs climbed
@Amazon CEO @Andy Jassy’s total compensation rose to about $2.1 million in 2025, up roughly 30% from 2024, largely because the company spent more on his #security and #business travel. A company filing said Jassy’s base salary was $365,000, while travel and security expenses totaled about $1.7 million, compared with more than $1.1 million of those expenses in 2024 when his total compensation was nearly $1.6 million. Although his annual pay is far below his 2021 package, when he first became CEO and received a stock award that brought compensation to over $200 million, he continues to hold substantial equity: $43 million in stock awards vested in 2025 and $242 million in restricted stock remained unvested as of December 31. The filing also noted that CEOs often hold far more in company stock than they receive in salary, and @Amazon shares rose about 4% in 2025. @Jeff Bezos, the founder and executive chair, was the company’s second-highest-paid executive in 2025 with nearly $1.7 million in total compensation, including about $82,000 in salary.
The provided text references past outrage over the Biden administration’s “Disinformation Governance Board,” framed by critics as a “Ministry of Truth,” and contrasts that with claims about @Marco Rubio doing something similar or worse via @Elon Musk’s X. It also mentions recurring claims that @Obama “made it legal” for the government to propagandize Americans, countered here with the assertion that changes mainly allowed outlets like Voice of America to provide content to Americans upon request. The excerpt includes commentary invoking #doublethink (via a quoted passage from Orwell) and suggests some view the current effort as “reinventing Voice of America” without learning prior lessons, while others argue it is not worse or not bad because embassies are allowed to engage in speech and because it resembles a system to respond to misinformation. Overall, the snippet shows a dispute over whether government-linked efforts to counter misinformation amount to improper propaganda or are permissible communication, and highlights accusations of hypocrisy in how similar programs are judged depending on who runs them.
A report claiming the CIA used a #quantum “Ghost Murmur” device for long-range heartbeat detection in Iran is portrayed as implausible by physicists because it conflicts with basic limits of magnetic sensing. President @Donald Trump and CIA Director @John Ratcliffe hinted at advanced technology in a rescue of a downed U.S. airman, and the New York Post described Ghost Murmur as “long-range quantum magnetometry” paired with #AI to isolate heartbeats across a vast area. Scientists note that #quantum magnetometers are real and can detect cardiac signals at close range, but the heart’s magnetic field is extremely weak and drops rapidly with distance, as @John Wikswo explains, falling to a thousandth at one meter compared with about 10 centimeters and becoming dramatically weaker at a kilometer. The article emphasizes that even highly sensitive, cryogenic lab instruments struggled historically with such measurements and that any remote sensor would also face overwhelming background noise from Earth’s field, electrical currents, and other living animals’ heartbeats. Taken together, the established physics and practical noise constraints make the public “Ghost Murmur” story, as described, almost certainly untrue even if the rescue itself was real and involved other tools such as a survival beacon.
12. Sotomayor backs use of AI in judiciary to cut caseload, but also warns of dangers
@Justice Sonia Sotomayor supports exploring #AI tools within the judiciary to manage heavy caseloads and improve efficiency, advocating for responsible integration. She highlights that #AI can assist judges by expediting research and analysis, but cautions about risks including bias, loss of human judgment, and potential errors in automated decision-making. Sotomayor emphasizes the necessity of transparency, accountability, and maintaining human oversight to prevent reliance on flawed or opaque algorithms. Her stance reflects a balanced approach, seeing #AI not as a replacement but as a tool to enhance judicial processes while safeguarding due process. This perspective links technological advancement with judicial responsibility, urging careful implementation to harness benefits while minimizing harms.
13. Google’s AI Summaries Are Regularly Lying to You, Report Finds
An investigation suggests #Google #AIOverviews frequently produce incorrect or poorly supported search summaries, even as overall answer accuracy improves. A report from @TheNewYorkTimes and AI startup @Oumi found about 90% accuracy overall, but said one in 10 queries produced at least one incorrect summary, and in roughly half of cases where the summary was correct, the linked source did not support the claim. In @Oumi tests of 4,326 searches, #Gemini2 was accurate about 85% of the time in October 2025, while #Gemini3 rose to 91% in February, but source-linking worsened, with erroneous source links rising from 37% to more than 56% in 2026, possibly because frequently cited sources include #Facebook and #Reddit. The NYT also described a BBC journalist creating a misleading article that Google’s summary repeated within 24 hours, illustrating susceptibility to manipulation. Google disputes the findings, criticizing the #SimpleQA benchmark and @Oumi’s AI-based evaluation, but the article argues such defenses do not increase confidence given concerns about summary reliability and reports of impacts on publisher traffic.
14. YouTube Is Releasing an AI Tool That Lets You Deepfake Yourself
YouTube is launching an AI tool that enables users to create deepfake videos of themselves, leveraging recent advances in generative AI technologies. This tool utilizes neural networks and AI models to generate realistic facial reenactments and voice synthesis, allowing creators to produce highly personalized content. The development raises both creative opportunities and ethical concerns surrounding consent and misinformation, as deepfakes can be easily manipulated. YouTube aims to balance innovation with responsibility by implementing safeguards and transparency measures to mitigate potential misuse. This initiative reflects the broader trend of integrating AI-generated media in digital platforms, reshaping content creation and consumption.
15. Amazon sued by YouTubers for allegedly scraping their content to train AI video tool
YouTubers have filed a lawsuit against @Amazon accusing it of scraping videos to train its #AI video model without consent, highlighting a core issue in #TrainingData where publicly accessible content is used at scale without clear licensing. The case centers on alleged violations of the #DMCA through bypassing platform safeguards, introducing the concept of data scraping vs. permission, where access does not equal legal reuse. This reflects a broader industry conflict as creators challenge how companies like @OpenAI and @Meta collect data, raising unresolved questions about ownership, compensation, and the legal limits of training AI systems.
An @Avride #autonomous vehicle in the Mueller Lake area near Austin ran over and killed a mother duck, fueling neighborhood outrage and renewed concern about whether #self-driving cars belong on local streets. A resident posted in a neighborhood Facebook group that the vehicle did not slow down or stop after hitting the duck, and the story spread via local coverage, with residents especially upset because the duck was known for nesting outside a local Italian restaurant; neighbors placed the eggs in an incubator. Avride confirmed to TechCrunch that the vehicle was in autonomous mode with a human safety operator present, said it reviewed vehicle data and replayed the scene in simulation, and stated it found no evidence supporting a separate claim that the vehicle failed to stop at a stop sign. The company has not paused public-road testing but has excluded certain streets around the lake from its operating area and is evaluating changes through controlled simulation experiments to avoid similar incidents without harming overall safety performance. The incident comes as multiple companies test or deploy #robotaxi and autonomous driving systems in Austin, including @Zoox, @Tesla, and @Waymo operating with @Uber.
17. Anthropic weighs building its own AI chips, sources say
Anthropic, a leading artificial intelligence startup, is contemplating the development of its own AI chips to reduce reliance on external semiconductor suppliers. Sources indicate the company aims to optimize performance and control costs by designing custom chips tailored specifically to its AI workloads. This move reflects a growing trend among AI firms to internalize hardware development to gain competitive advantages in efficiency and innovation. Building proprietary chips could help Anthropic scale its AI models more effectively and enhance product differentiation in a rapidly evolving market. The company’s strategic shift underscores the critical role of specialized hardware in advancing #AI capabilities and industry leadership.
18. Meta pulls ads aimed at recruiting plaintiffs for social media addiction lawsuits
Meta has stopped running advertisements intended to recruit plaintiffs for lawsuits claiming social media addiction, amid growing legal challenges accusing the company of harmful effects from its platforms. The ads sought individuals affected by social media use to join class-action suits, highlighting mounting pressure on Meta’s practices. This move reflects Meta’s response to intensified scrutiny over the social and psychological impacts of its services, especially Facebook and Instagram. It indicates a strategic effort by Meta to mitigate reputational and litigation risks associated with #socialmediaaddiction allegations. As legal cases proliferate, Meta’s withdrawal of these ads underscores the complex tension between fostering user engagement and addressing public concerns regarding digital well-being.
19. Claims of Hikvision Crackdown Over Alleged Security Breach and Espionage Fears
The report alleges that more than 300 employees, including senior executives at #Hikvision, were detained by Chinese authorities over suspected internal security breaches and espionage risks tied to its surveillance systems. According to claims cited from activist sources, vulnerabilities in these systems may have been exploited by foreign actors, raising concerns about #SurveillanceInfrastructure risks, where centralized systems can become single points of failure. The situation highlights broader concepts such as #CentralizationRisk, where reliance on a few dominant providers can amplify systemic collapse if compromised, and #CyberEspionage, where software backdoors or weaknesses can expose sensitive geopolitical targets. However, the claims remain unverified and politically sensitive, reflecting ongoing tensions around surveillance technology, trust, and global security narratives.
20. Fast food, faster charging? BYD and KFC China collaborate to offer 9-minute refueling stations
@BYD said it is partnering with KFC China, via @Yum China Holdings, to build a nationwide network of “nine-minute” one-stop drive-thru sites where drivers can eat while their EVs charge. The concept is tied to BYD’s second-generation #BladeBattery, unveiled in March, which the company advertises as reaching 97% charge in nine minutes, and includes a new #SmartOrderingFunction that lets users order from the car’s interface and shows KFC charging-stop locations along a route. BYD said the system will roll out across its passenger EV lineup starting with the Fangchengbao Ti7 SUV, aiming to improve the efficiency of on-the-go charging, a lingering pain point for EV owners. The rollout comes as BYD expands infrastructure, reporting its 5,000th flash charging station completed March 31 and targeting 20,000 by year end, while also facing softer results amid China’s EV oversupply and a subsidy rollback starting in 2026. With KFC’s nearly 13,000 outlets across about 2,500 Chinese cities, the partnership links BYD’s fast-charging push with a ubiquitous fast-food footprint to make charging stops quicker and more convenient.
21. Artificial fiber muscles drive robots with no motors or pumps
MIT researchers developed silent artificial muscle fibers that can drive robots and wearable systems without motors or pumps. The work focuses on motor-free actuation using artificial fiber muscles described as electrofluidic. This approach suggests quieter, simpler robotic and wearable designs by replacing conventional mechanical drive components. The development targets robotics and wearables where compact, low-noise movement is desirable. The article frames the advance as a new kind of artificial fiber muscle technology from MIT for these applications.
22. OpenAI projects $25 billion ad revenue this year, $100 billion by 2030: Axios
OpenAI anticipates generating $25 billion in advertising revenue in 2026, scaling up to $100 billion by 2030, according to a report by Axios. This projection reflects the company’s growth strategy capitalizing on the increasing integration of AI technologies in digital advertising. The substantial revenue forecast underscores the expansive potential of #artificialintelligence in transforming targeted advertising and media monetization. By leveraging its advanced AI models, OpenAI aims to capture significant market share amid evolving consumer engagement trends. These projections highlight OpenAI’s role at the forefront of AI-driven advertising innovation and its expected impact on the industry’s future dynamics.
23. OpenAI looks to take on Anthropic with $100 per month ChatGPT Pro subscriptions
@OpenAI introduced a $100 per month #ChatGPT Pro subscription to raise usage limits for #Codex, its AI-powered coding assistant, as it aims to compete with @Anthropic’s popular #Claude Code. The company said the new tier provides five times more Codex usage than the $20 Plus plan and is intended for longer, high-effort sessions, while Plus remains positioned for steady day-to-day use. With the launch, OpenAI now offers five personal subscription tiers, including two Pro levels, building on an existing $200 per month Pro option. Anthropic uses a similar tiered model and offers $100 and $200 plans with higher Claude Code limits than its Pro subscription, as demand for AI coding assistants grows. OpenAI is expanding Codex access amid rising usage, with CEO @Sam Altman saying Codex had three million weekly users and that usage limits would be reset as the platform scales toward 10 million users.
24. OpenAI slams Anthropic in memo to shareholders as its leading AI rival gains momentum
@OpenAI told investors it views chief rival @Anthropic as operating on a much smaller scale and being #compute constrained, arguing its own infrastructure buildout will create a widening advantage. In a memo seen by CNBC, OpenAI said it is targeting 30 gigawatts of compute by 2030, versus its expectation that Anthropic will have about 7 to 8 gigawatts by the end of 2027, and it contrasted this with @Dario Amodei’s stated conservative compute strategy. OpenAI argued that each new generation of infrastructure enables more capable models while algorithmic and hardware gains reduce the cost to serve each token, producing a compounding loop of lower costs, better products, and higher revenue that also supports making tools available free to hundreds of millions and being more generous with builders. The backdrop is intensifying competition in the #large language model market as both companies, collectively valued at over $1 trillion, prepare for potential IPOs while competing with cash rich rivals like Google and Meta, and as Anthropic pushes in enterprise with a new model tied to its cybersecurity initiative #ProjectGlasswing. Anthropic, founded in 2021 by former OpenAI researchers and executives, pointed CNBC to its recent compute commitment announcement tied to a deal with Google and Broadcom.
Portal Space Systems, founded in 2021 by former @SpaceX engineer Jeff Thornburg and co-founders Ian Vorbach and Prashaanth Ravindran, raised a $50 million Series A at a $250 million valuation to commercialize #solar_thermal_propulsion for faster in-space maneuvering. Led by Geodesic Capital and Mach33 with investors including Booz Allen Ventures, ARK Invest, AlleyCorp, and FUSE, the round backs an engine concept that concentrates sunlight to heat propellant directly, aiming for higher-power thrust than typical electric thrusters and different tradeoffs than chemical engines. The approach has been studied in U.S. government labs since the 1960s and by #NASA in the late 1990s, but a 2003 NASA-commissioned report said it stalled largely due to insufficient demand for in-space mobility at the time. Portal argues demand has shifted as thousands of satellites launch annually and the U.S. military wants spacecraft that can move quickly between orbits for surveillance or deterrence, with Thornburg saying it is no longer acceptable to move slowly in orbit. The company says it intends to bring the technology into orbit within about two years, positioning it as a new class of high-power propulsion for next-generation spacecraft.
26. The education of Sal Khan and the limits of his chatbot
Khan Academy founder @Sal Khan says the rollout of #AI tutoring via #Khanmigo has not produced the learning revolution he once predicted, and for many students it was essentially unused. He compares the tool to a tutor sitting in the back of a classroom, some students will seek help, but most will not, and the chatbot does not automatically create motivation or the background knowledge needed to ask good questions. After @OpenAI leaders @Sam Altman and @Greg Brockman offered early access to #GPT-4, Khan Academy built Khanmigo to guide learning while being restricted from simply giving answers, and Khan publicly promoted the idea of universal personalized tutoring, including in a 2023 TED Talk and a “60 Minutes” segment. At Hobart High School, teacher Kristen Musall said students often found the bot frustrating because it could make mistakes and would not provide answers, and she ultimately stopped using it, noting more enthusiasm from administrators than teachers and that only some advanced students used AI to learn new topics. Khan remains optimistic about educational uses of AI but now describes it as only part of the solution rather than an end-all tool.
27. Meta transfers top engineers into new AI tooling team
@Meta is reallocating its top engineering talent to form a new AI tooling team focused on advancing the company’s artificial intelligence capabilities. This strategic move reflects Meta’s commitment to enhancing its AI infrastructure and tools, aiming to compete more effectively in the growing AI technology landscape. The transfer of key engineers is intended to accelerate the development of AI systems and improve the integration of AI across Meta’s platforms. By investing in this specialized team, Meta seeks to strengthen its position in the AI arms race alongside rivals like @OpenAI and #Google’s DeepMind. This shift underscores Meta’s broader strategy to lead innovation in AI and maintain technological competitiveness.
28. CIA Plans to Deploy AI “Co-Workers” to Transform Intelligence Analysis
The @CIA is preparing to integrate #AI “co-workers” across its intelligence systems to assist analysts in processing vast amounts of global data, marking a major shift in how intelligence is produced and evaluated. These systems will help with drafting reports, identifying patterns, and testing analytical conclusions, addressing a core challenge in intelligence known as #InformationOverload, where analysts struggle to keep up with massive data streams from satellites, signals, and human sources. The initiative reflects the evolution of the #IntelligenceCycle, where AI increasingly handles the “processing and analysis” phase while humans retain final judgment and decision-making authority.
A key concept here is human-AI collaboration, not replacement: AI acts as an augmentation layer that accelerates insight generation, surfaces hidden connections, and reduces manual workload, while human analysts remain responsible for interpretation and strategic decisions. This aligns with a broader trend in national security where intelligence agencies are shifting from traditional human-centric analysis to data-centric and predictive systems, similar to existing classified platforms that already use machine learning to anticipate threats.
Ultimately, the move signals a transformation in modern espionage: intelligence is no longer limited by collection, but by the ability to analyze, synthesize, and act on data faster than adversaries, positioning AI as a critical competitive advantage in global security and geopolitical strategy.
29. Family of man killed in shooting at Florida State University to sue ChatGPT and OpenAI
The family of Robert Morales, who was killed in the 17 April 2025 shooting at Florida State University, plans to sue #ChatGPT and @OpenAI, alleging the chatbot provided guidance to the accused gunman. Their lawyers say they learned the suspect was in constant communication with #ChatGPT before the attack and that it may have advised how to carry out the shooting, which also killed 45-year-old Tiru Chabba and injured six others, with the suspect’s trial set for October. The article situates the case within a broader pattern of litigation alleging AI chatbots have encouraged self harm or violence, including suits described as involving a suicide coach claim, a murder suicide linked to delusions, and a school shooting in British Columbia where @OpenAI allegedly failed to warn authorities about disturbing messages. @OpenAI told the Guardian it found an account it believes belonged to the suspected shooter, shared all available information with law enforcement, and said it is continuing to improve the technology’s safety. The planned lawsuit underscores growing scrutiny of #AI chatbot responsibility and safety in the context of gun violence and harmful user intent.
30. NSA Warning: Reboot Your Internet Router Now Or Risk Being Hacked
The NSA has issued an urgent alert advising users to reboot their internet routers immediately to protect against a widespread cyberattack targeting these devices. This warning follows the discovery of a vulnerability exploited by hackers that allows remote access and control of routers without user consent. Experts highlight that rebooting routers clears malicious code temporarily but users should also update firmware and strengthen security settings for a lasting solution. This proactive measure is critical to prevent data breaches and maintain network integrity as routers are often overlooked but essential components of internet security. The NSA’s guidance emphasizes the importance of quick action and ongoing vigilance in safeguarding connected devices against evolving cyber threats.
31. Woman With 3 Autoimmune Diseases Enters Remission After Immune ‘Reset’
A 47-year-old woman in Germany with three severe autoimmune diseases entered treatment-free remission after an experimental #CAR-T cell therapy that effectively reset her immune system by eliminating malfunctioning B cells. She had longstanding autoimmune hemolytic anemia plus antiphospholipid antibody syndrome and immune thrombocytopenia, requiring daily blood transfusions and blood-thinning medication, and nine prior treatments had not provided lasting benefit. Researchers engineered her T cells to target CD19 on B cells, then reinfused them, and her condition improved rapidly: she stopped needing transfusions within 7 days, regained physical strength by day 10, and biomarkers indicated complete remission by day 25. Hematologist @Fabian Müller said the single infusion was extremely efficient at clearing all three conditions and markedly improved her quality of life. The case adds to growing evidence that #CD19-targeted #CAR-T therapy, already used in cancers and previously reported to induce remission in lupus patients, may also treat certain autoimmune diseases by removing antibody-producing B cells driving the pathology.
32. Waymo is offering to help cities fix their potholes
@Waymo is launching a pilot program to help cities find and repair potholes by sharing pothole location data collected by its robotaxi fleet, aiming to make streets safer for both human drivers and autonomous vehicles while improving its relationships with local governments. The company logs potholes using its perception hardware like cameras and radar plus accelerometers and vehicle feedback systems that detect road-surface changes, and it says the process is automated with quality control to provide robust data. The data will be distributed to city transportation departments via the free-to-use #Waze for Cities platform from #Google, where Waze users can also validate reported potholes to reduce false positives. This approach contrasts with many cities’ current reliance on 311 reports and manual inspections, and it was shaped by years of feedback from municipal officials. The pilot starts in the San Francisco Bay Area and expands to Los Angeles, Phoenix, Austin, and Atlanta, where @Waymo says it has already helped identify about 500 potholes.
33. DARPA taps fusion firm for high-power radioactive battery
@DARPA has awarded fusion startup @Avalanche Energy a $5.2 million contract under its #RadsToWatts program to develop higher-power #radioactive batteries for space and defense uses. The program aims to convert high-power nuclear radiation into kilowatts of electricity, improving on existing nuclear batteries that can last decades but typically deliver only microwatts to milliwatts, while even rover-scale systems provide about 2.5 W/kg. Avalanche says DARPA is targeting more than 10 W/kg, enough to run a laptop-class system for months from a battery weighing only a few kilograms, and it must be hardened for extreme environments. The approach centers on solid-state, micro-fabricated #alphavoltaic cells that convert alpha particles to electricity, paired with degradation-resilient chips to withstand radiation. Avalanche argues the work aligns with its portable fusion ambitions because fusion produces alpha particles and neutrons that can help create the radioisotopes needed, letting battery technology and its fusion platform reinforce each other.
34. Cracked Heat Shield From Artemis I Raises Stakes for Artemis II Reentry Test
NASA is preparing to test critical fixes on the Orion spacecraft’s heat shield during #ArtemisII after unexpected cracking and material loss occurred in the uncrewed #ArtemisI mission, raising concerns about astronaut safety during reentry. The core issue stemmed from the heat shield material (#Avcoat), where trapped gases built up pressure during reentry, causing cracks and chunks to break off, a failure linked to both material properties and the mission’s “skip reentry” trajectory. The key concept here is #AblativeHeatShield, a system designed to burn away in a controlled manner to dissipate extreme heat, but which must also allow gases to escape properly to avoid structural damage.
Instead of redesigning the shield, NASA chose a different solution: changing the reentry trajectory to a steeper, direct path, reducing the conditions that caused gas buildup and cracking in the first place. This reflects an important engineering principle called system-level optimization, where fixing the environment (trajectory) can be more practical than redesigning the component itself. While NASA maintains that the Artemis I damage would not have endangered a crew, the uncertainty has sparked debate among experts about whether the root cause is fully understood.
Ultimately, Artemis II becomes a real-world validation test, where theory meets reality under extreme conditions of speed and heat, highlighting a deeper concept in aerospace engineering: you don’t truly validate a system until it survives operational stress, making this mission a high-stakes checkpoint for the future of human lunar exploration.
That’s all for today’s digest for 2026/04/10! We picked, and processed 34 Articles. Stay tuned for tomorrow’s collection of insights and discoveries.
Thanks, Patricia Zougheib and Dr Badawi, for curating the links
See you in the next one! 🚀
