Welcome to today’s curated collection of interesting links and insights for 2025/10/28. Our Hand-picked, AI-optimized system has processed and summarized 28 articles from all over the internet to bring you the latest technology news.
As previously aired🔴LIVE on Clubhouse, Chatter Social, and TikTok.
Also available as a #Podcast on Apple 📻, Spotify🛜, Anghami, and Amazon🎧 or anywhere else you listen to podcasts.
1. OpenAI to give free 1-year ChatGPT Go subscription to India users amid big push
OpenAI is offering Indian users a free one-year subscription to ChatGPT Go, its premium AI service, as part of a strategic effort to expand its presence in the growing AI market in India. This move aims to attract a broader user base by letting individuals experience enhanced features such as faster response times and access to the latest GPT-4-turbo model. The initiative underscores OpenAI’s focus on India, a country with a rapidly increasing demand for AI technologies and digital services. By providing free access to advanced AI tools, OpenAI intends to build loyalty and encourage adoption ahead of competitors. This approach aligns with the company’s broader mission to democratize AI and foster innovation globally, starting with key markets like India.
2. The AI Audio Processing Revolution: Transforming Sound in Real Time
AI-driven audio processing is rapidly transforming the way sound is captured, enhanced, and manipulated in real time. Technologies like neural networks and deep learning enable applications ranging from noise cancellation in consumer devices to sophisticated music production and adaptive sound environments. Pioneers such as @OpenAI and #GoogleDeepMind are developing models that understand and generate audio content with unprecedented accuracy, improving clarity and creating new creative possibilities. This revolution addresses challenges in latency and computational load, making high-quality audio processing accessible on mobile and embedded systems. As AI audio tools evolve, they are set to redefine industries like entertainment, communication, and accessibility by offering smarter, more intuitive audio experiences.
3. PayPal Users Warned: Do Not Pay, Do Not Phone As Attackers Strike
PayPal users are being targeted by attackers urging them not to pay or contact customer service directly, as scammers intensify efforts to exploit the platform. Reports highlight a surge in phishing attempts and fraudulent messages impersonating PayPal to coerce victims into unauthorized payments or divulging sensitive information. These tactics emphasize social engineering and deception to bypass PayPal’s security measures. Users are advised to verify communications through official channels and remain vigilant against unsolicited requests. Maintaining awareness of such scams is critical to protecting personal and financial data on widely used digital payment systems like PayPal.
4. Elon Musk’s Grokipedia Pushes Far-Right Talking Points
@ElonMusk’s xAI project Grokipedia launches as an AI generated alternative to Wikipedia, but WIRED’s review finds most pages resemble Wikipedia in tone while some entries push conservative viewpoints and propagate historical inaccuracies. The site initially blocked, and once accessible it contained AI generated entries that challenge mainstream media, praise conservative perspectives, and include false claims such as that porn worsened the HIV/AIDS epidemic and that transgender topics are framed in denigrating terms. One entry on slavery discusses ideological justifications and critiques the 1619 Project, while a search for gay marriage yielded a suggestion for gay pornography and an HIV AIDS claim tied to porn use. The platform’s coverage of Trump and other topics rails against left leaning media, and Musk’s own criticisms of WIRED are incorporated into Grokipedia’s narrative. The combination of biased entries and misinformation under a banner of AI generated knowledge raises questions about the integrity and goals of #Grokipedia and #xAI, and about the risk of distorted information spreading on #AI powered knowledge platforms.
5. Elon Musk’s Grokipedia contains copied Wikipedia pages
Grokipedia, @ElonMusk’s Wikipedia-like site, is live but shows substantial overlap with #Wikipedia, including pages labeled as ‘adapted’ from Wikipedia. On the MacBook Air page—and also on pages for the PlayStation 5 and the Lincoln Mark VIII—the site carries the notice ‘The content is adapted from #Wikipedia, licensed under CC BY-SA 4.0,’ and some entries are almost word-for-word identical to their Wikipedia counterparts. A spokesperson for the Wikimedia Foundation notes Grokipedia still relies on Wikipedia, highlighting concerns that even ‘fact-checked’ entries can reflect AI-generated inaccuracies. The Verge notes this is an early, v0.1 version, with Grokipedia listing over 885,000 articles versus Wikipedia’s roughly 7 million English pages. @ElonMusk previously promised that Grokipedia would be a ‘massive improvement’ over Wikipedia, a claim that remains to be tested as the overlap persists.
6. Ransomware profits drop as victims stop paying hackers
Ransomware profits have declined significantly as victims increasingly refuse to pay hackers, reducing cybercriminal incentives. Reports show a marked drop in ransom payments due to improved cybersecurity practices and greater awareness among organizations. This shift is undermining the ransomware economy and forcing threat actors to adapt their tactics or target less prepared victims. As more companies hesitate to comply with demands, the profitability of ransomware attacks diminishes, potentially altering the cybercrime landscape. The trend highlights the critical impact of resilience and refusal to pay in combating cyber extortion.
7. Signal CEO Meredith Whittaker opens up on AWS dependency and its impact on privacy
Signal CEO Meredith Whittaker has addressed the challenges and risks related to the app’s reliance on Amazon Web Services (AWS) for infrastructure. Whittaker points out that while AWS provides critical backend support, this dependency raises concerns about #dataprivacy and control given AWS’s prominence in the cloud market. She highlights the tension between using major cloud providers to scale securely and the desire to minimize exposure to centralized platforms that may compromise user privacy. This frank perspective sheds light on the ongoing balancing act faced by privacy-focused applications like Signal in leveraging powerful but potentially conflicting technological resources. The conversation underscores the broader debate on cloud dependency tensions within the #tech and #privacy communities.
8. Chemists develop novel antibiotic against drug-resistant bacteria
A team of chemists has developed a novel antibiotic to combat drug-resistant bacteria, addressing the critical challenge of antibiotic resistance. The antibiotic operates through a unique mechanism that bypasses traditional pathways targeted by existing drugs, as demonstrated in laboratory tests against multiple resistant bacterial strains. This innovation could significantly reduce the threat posed by drug-resistant infections, which are a growing global health concern. By offering an effective alternative to current treatments, the new antibiotic exemplifies the potential of chemistry-driven solutions in medical science. This advancement underscores the importance of continued research into #antibiotics and #drugresistance to safeguard public health.
9. Why Bubbles Can Keep Inflating in Plain Sight
Bubbles in financial markets can continue growing despite widespread awareness because investors often justify rising prices with new narratives or innovations that suggest fundamentals have changed. The article highlights that bubbles are driven by human psychology, particularly the fear of missing out and herd behavior, rather than purely economic metrics. It discusses how prominent figures like @AlanGreenspan acknowledged the challenge of identifying bubbles while they inflate, and how markets tend to discount risks in bullish phases. This dynamic allows speculative manias to persist, as participants rationalize higher valuations by pointing to disruptive technologies or policy changes. Ultimately, the article links bubble persistence to the complex interplay of optimism, innovation, and market psychology.
10. OpenAI tightens ChatGPT’s responses on suicide and self-harm
OpenAI has updated ChatGPT’s programming to better handle conversations about suicide and self-harm, aiming to provide safer and more supportive responses. The changes include refined moderation tools and improved training data to ensure the AI encourages users to seek professional help and avoids triggering content. These measures address concerns about AI’s role in managing sensitive mental health topics and the potential risks of misinformation or harmful suggestions. By enhancing the chatbot’s responses, OpenAI seeks to foster a more responsible interaction environment while balancing user needs for empathy and guidance. This development highlights ongoing efforts in AI ethics and safety, reinforcing the importance of mental health considerations in technology design.
11. US government uses Halo images in a call to ‘destroy’ immigration, Microsoft declines to comment
The US government has co-opted imagery from the Halo franchise to promote deportation efforts, continuing a pattern of using copyrighted material for propaganda. A White House post showed an AI-generated image of @Donald Trump as Master Chief, followed hours later by the Department of Homeland Security posting a Halo image with the message ‘Destroy the Flood’ and a link to the ICE recruitment page. The episode also involved GameStop spreading the faux diplomatic post and retweeting related imagery, while a Microsoft representative declined to comment on the matter. The piece warns that dehumanizing immigrants by likening them to alien invaders echoes propaganda from history, citing references to the Holocaust and language used in Rwanda. Among fans, discourse in Halo communities contrasted with official messaging, underscoring tensions over political use of game imagery and the lack of a public Microsoft response.
12. Sleeping Tesla Driver Blames Autopilot for Rear-Ending Police Car
A sleeping Tesla driver in Illinois says Autopilot was in control when the car rear-ended a police cruiser, a claim echoed by the @South Barrington Police Department in a Facebook post. The piece notes Autopilot is a limited, hands-free system that requires the driver to stay attentive, and it cautions that technology can assist but does not replace human judgment behind the wheel. Details are sparse: the department’s photos do not show the car’s year or model, and the driver’s account cannot be independently verified, with the author pointing to past incidents where driver stories proved unreliable, including a fire truck incident. The analysis suggests it is unclear whether Autopilot was active at impact, emphasizing the ongoing uncertainty and the responsibility of occupants for safe operation. No officers were injured, and authorities remind drivers that accountability remains with the occupants even if the machine is driving.
13. Vercel trained an AI agent on its best salesperson. Then it cut the 10-person team down to 1.
Vercel trained an AI agent using the work of its top salesperson to automate core sales tasks. The approach is evidenced by the description that the AI was trained on the top rep’s output and that the project reduced a 10-person sales team to a single role. This suggests potential efficiency gains from automation, alongside the need to reallocate human workers to more creative tasks and maintain appropriate oversight. The case illustrates how enterprise AI can reshape sales teams and productivity, signaling a shift toward AI-augmented workflows in tech companies. #AI #sales #workforce #automation
14. Waymo co-CEO on robotaxi vandalism: ‘We’re not standing for it’ | TechCrunch
Waymo’s @Tekedra Mawakana says vandalism of robotaxi vehicles is ‘not okay’ and that the company will work with law enforcement to pursue those responsible, regardless of whose fleet is targeted. She cites recent vandalism incidents, including a weekend attack in Los Angeles, and stresses that damaging property is unacceptable and that Waymo ‘is not standing for it.’ The co-CEO also reiterates Waymo’s position on privacy, refusing government footage requests that are overly broad or unlawful, while stressing ongoing cooperation with first responders. To address public perception, Waymo is pursuing community engagement, such as adding local artists’ designs to some cars to make the fleet feel more part of the community. The broader message is that safety, accountability, and privacy must be balanced as the company confronts vandalism and surveillance concerns.
15. OpenAI says U.S. needs more power to stay ahead of China in AI: ‘Electrons are the new oil’
OpenAI argues the United States must massively expand its energy capacity to stay ahead of #China in #AI, framing electricity as a strategic asset rather than a utility. In a blog post and an 11-page submission to the White House Office of Science and Technology Policy (#OSTP), it urged building 100 gigawatts of new energy capacity each year to power sprawling data centers. It notes that #China added 429 GW last year while the U.S. added 51 GW, creating an electron gap and stating that electrons are the new oil. The plan comes as OpenAI’s infrastructure buildouts raise concerns about grid strain, a point echoed by @Sam Altman. If the U.S. fails to scale energy capacity, leadership in #AI could be at risk, linking energy policy to national AI leadership.
16. U.S. Department of Energy forms $1 billion supercomputer and AI partnership with AMD: Reuters
The U.S. Department of Energy and @AMD have formed a $1 billion partnership to build two supercomputers that will tackle large scientific problems from nuclear power and fusion energy to cancer treatments and national security, as Energy Secretary @Chris Wright and @Lisa Su told Reuters. The first system, Lux, will come online within six months and is built around AMD’s MI355X AI chips with CPUs and networking, co-developed by @AMD, @HPE, Oracle Cloud Infrastructure, and ORNL. Lux is expected to deliver about three times the AI capacity of current supercomputers, per ORNL Director Stephen Streiffer. The second system, Discovery, based on AMD’s MI430 AI chips tuned for high-performance computing, will be delivered in 2028 and ready for operations in 2029, designed by ORNL, HPE, and AMD. DOE will host the computers, the companies will provide the machines and capital spending, and both sides will share computing power; officials say this could be the first of many such public-private partnerships with private industry and DOE labs across the country, with Wright predicting faster fusion progress and tangible advances in cancer treatment and drug discovery in the years ahead. #AI #supercomputers #fusion #drugdiscovery #nuclearsecurity
17. Qualcomm stock jumps 11% as company enters AI chip race, taking on Nvidia, AMD
Qualcomm is entering the AI data-center market with its AI200 and AI250 chips and rack-scale servers, signaling a broader push beyond smartphones to compete with @Nvidia and @AMD in the AI hardware race #AI #dataCenter. The AI200 is both an AI accelerator and the server rack it slots into, while the AI250 is a next-generation accelerator with the AI250 promising 10x the memory bandwidth of the AI200, and Qualcomm plans to add a third chip in 2028 and future generations on an annual cadence #HexagonNPU #inference. Chips are designed for AI inference rather than training, and customers won’t need to buy Qualcomm servers to access them, as they can purchase individual chips or portions of the server offerings, potentially including rivals like @Nvidia and @AMD as partners or competitors #server. This effort fits Qualcomm’s broader aim to diversify from handset licensing, with Q3 revenue at $10.4B and $6.3B from handsets, while data-center revenue remains unreported for now. Past attempts such as the Centriq 2400 project faltered amid stiff competition and legal issues, underscoring the uphill path to meaningfully scale in cloud AI hardware even as Qualcomm also markets an AI 100 Ultra card today.
18. New image-generating AIs are being used for fake expense reports
AI-generated receipts are increasingly used to submit fake expense reports, driven by new image-generation models from @OpenAI and @Google that produce convincing submissions #image-generation #AI. Evidence shows fake AI receipts accounted for about 14% of fraudulent documents in September, while Ramp flagged over $1 million in fraudulent invoices in 90 days, and about 30% of US and UK financial professionals reported a rise in falsified receipts after GPT-4o’s launch. The realism of these receipts has pushed expense platforms to rely on AI-driven detection that scans image metadata and cross-checks trip details and server data, though users can defeat metadata by taking photos or screenshots. Previously, fraud required photo-editing skills or paid services; now free image-generation tools let anyone generate receipts in seconds, prompting CFOs to say that AI-assisted falsifications are already occurring, with nearly 70 percent suspecting and about 10 percent certain it has happened. In response, platforms like SAP Concur perform millions of AI-enabled compliance checks monthly, and industry voices such as Rydoo advocate stronger AI-based detection as the threat grows.
19. Amazon targets many as 30,000 corporate job cuts, sources say
Amazon plans to cut about 30,000 corporate jobs globally, according to sources familiar with the matter. The reductions target corporate roles to align operating expenses with lower business growth and improve efficiency amid economic pressures and slowing consumer spending. This significant workforce reduction is part of Amazon’s ongoing effort to optimize costs following previous rounds of layoffs affecting thousands of employees. Although specific departments and the timeline remain undisclosed, the cuts represent one of the largest corporate job reductions in the company’s history. This move underscores Amazon’s strategic shift toward leaner operations to maintain competitiveness in a challenging market environment.
20. Ultra-HD televisions not noticeably better for typical viewer, scientists say
Ultra-HD TVs delivering #4K or #8K resolution do not provide a noticeable sharpness advantage for the average living room when compared with a similarly sized 2K display. In a study by researchers from @Cambridge and @Meta, a 27-inch 4K monitor was moved to different viewing distances and shown to 18 participants who judged line patterns against a grey reference to determine the #resolution limit. They found the eye can resolve about 94 #PPD for greyscale and 89 #PPD for red/green patterns, with yellow/violet at only 53 #PPD, indicating many extra pixels are not perceptible. A separate test with white text on black showed the same perceptual limit, reinforcing that higher resolution beyond what the eye can discern yields no visible benefit #text. The team released a chart mapping screen size and viewing distance to the nearest standard resolution and a free online calculator to help people decide whether upgrading would improve what they see, implying upgrading beyond 2K is often wasteful #4K #8K.
21. Hundreds of People With Top Secret Clearance Exposed by House Democrats’ Website
House Democrats inadvertently exposed the names, hometowns, and social security numbers of hundreds of individuals holding top secret security clearances on a public website. The information was found in documents related to a 2019 House Oversight and Reform Committee investigation, which were posted online and accessible without restrictions. This exposure raises significant security concerns, as the data could be exploited for identity theft or espionage. The incident highlights vulnerabilities in how sensitive government information is managed and the need for stricter controls. It serves as a warning about the risks of mishandling classified or sensitive personal data on official government platforms.
22. Apple will bump up the iPhone 18 to 12GB of LPDDR5X RAM
The @Apple iPhone 18 is expected to arrive in Spring 2027 with 12GB of LPDDR5X RAM, up from 8GB on the base iPhone 17 and matching the 12GB found in the iPhone 17 Pro and Pro Max and the iPhone Air. Industry sources say Apple has asked @Samsung to boost RAM supply for the next iPhone lineup, with Samsung currently producing LPDDR5X RAM in 12GB and 16GB modules #RAM. The 12GB base would enable on-device #AI model execution that is difficult with 8GB. The report also notes a staggered release cadence from 2026, with Pro models launching in Fall 2026 and the base iPhone 18 following in Spring 2027, reflecting Apple’s supply strategy #releaseCadence.
23. New ChatGPT-Atlas Browser Exploit Lets Hackers Bypass Same-Origin Policy, Steal Data
A newly discovered security vulnerability in the ChatGPT-Atlas browser engine allows attackers to bypass the same-origin policy and steal sensitive data from users. This exploit enables malicious actors to access cross-origin content, compromising user privacy and security. Researchers have demonstrated that this flaw can be leveraged to execute cross-site scripting (XSS) attacks, potentially affecting numerous web applications relying on ChatGPT-Atlas. The breach highlights the challenges in securing AI-powered browser engines amidst increasing integration of machine learning technologies in web environments. Users and developers are urged to apply patches and adopt stringent security measures to mitigate risks posed by this critical exploit.
24. Sora is showing us how broken deepfake detection is
Sora demonstrates that even the best tagging systems for AI-generated content, like #C2PA Content Credentials embedded via the #C2PA framework, are not reliably visible or effective in everyday viewing, exposing how broken deepfake detection is. The platform can produce convincing videos featuring public figures such as @MartinLutherKingJr, @MichaelJackson, and @BryanCranston, as well as copyrighted characters like @SpongeBob and @Pikachu, including instances where voices or actions are misrepresented or used for harmful purposes. C2PA metadata is embedded in Sora clips, but you’d rarely notice it as the labeling on platforms like Instagram, TikTok, and YouTube remains barely visible or easy to miss, with many posts lacking clear provenance. @OpenAI sits on the C2PA steering committee, but the system only works if it’s adopted at every step of creation and posting, which most platforms fail to do, leaving viewers without reliable cues that content is AI-generated. In short, Sora exposes the gap between technological labeling ambitions and real-world practice, highlighting the need for stronger, more visible protections and universal adoption of provenance signals to restore trust in online media.
25. Italian spyware vendor linked to Chrome zero-day attacks
Italian cybersecurity company RCS Lab has been linked to zero-day Chrome vulnerabilities exploited in recent cyberattacks, highlighting concerns about state-sponsored spyware vendors using vulnerabilities to target dissidents. Google discovered these campaign attacks involved exploiting security flaws affecting multiple platforms, and promptly patched the Chrome zero-day vulnerability. RCS Lab’s spyware, known for surveillance capabilities, was observed being used to deliver malicious payloads through the Chrome exploit chain. This connection illustrates how commercial spyware developers are actively leveraging 0-day vulnerabilities to maintain access in high-profile targets, raising significant privacy and security implications. The revelation prompts the need for tighter scrutiny of spyware companies and enhanced collaboration between browser vendors and security researchers to defend against evolving threat actors.
26. Mushrooms could inspire memory chips of the future
Researchers have explored how the fibrous networks in mushroom roots called mycelium can inspire new designs for memory chips. By studying the structural complexity and information processing capabilities of mycelium, scientists identified parallels between natural fungal systems and electronic circuits. This comparison suggests that integrating principles from biology, specifically the self-organizing and adaptive features of mycelium, could lead to more efficient and resilient memory technologies. The research highlights a potential interdisciplinary approach combining #mycology and #microelectronics to innovate future computing devices. These insights may contribute to the development of memory chips that better mimic natural processes, fostering advancements in both bio-inspired engineering and information technology.
27. Google and NextEra to revive major Iowa nuclear facility as AI energy demand surges
@Google and @NextEra Energy plan to restart Iowa’s Duane Arnold Energy Center by 2029 to provide 24/7 carbon-free power for Google’s cloud and #AI operations, aiming to bolster local grid reliability. The 615-MW plant, which closed in 2020, could begin operating in early 2029 pending regulatory approval, with Google purchasing the power to support its expanding infrastructure. This move reflects a broader wave of tech-nuclear deals as U.S. electricity demand climbs due to #AI and #data-centers growth, signaling an industry push to secure reliable, low-carbon energy amid competition from natural gas and renewables. The project aligns with other partnerships like @Microsoft and @Constellation Energy and @Oracle’s data center initiatives powered by small reactors, illustrating nuclear energy as a potential path to sustain the AI-driven economy while supporting grid stability and jobs.
28. Space Startup Captures Chinese Mystery Satellite Before Fiery Reentry
XJY-7, a Chinese payload launched in December 2020 on the Long March 8, remained in geostationary orbit for nearly five years with an undisclosed purpose until Australian firm @HEO used a network of imaging satellites to capture it and build a 3D rendering that reveals a large dish antenna and two fixed solar panels. Using high-frequency Non-Earth Imaging, HEO produced multiple-angle views and verified XJY-7 as a SAR satellite capable of Earth observation, needing full-body rotation to charge. The satellite reentered Earth’s atmosphere on October 16 over Tenerife, with tracker Marco Langbroek reporting a bright, slow fireball and sonic booms; he identified the event as artificial space debris. The piece highlights the growing power of satellite-on-satellite imaging by firms like @HEO and @Maxar to reveal hidden details about China’s space hardware, and notes related activity such as #Shijian-25 in-orbit refueling tests as context for rapid tech development. It underscores how imaging networks are reshaping transparency around otherwise secret space programs and the pace of China’s space tech evolution.
That’s all for today’s digest for 2025/10/28! We picked, and processed 28 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! 🚀