Welcome to today’s curated collection of interesting links and insights for 2025/09/29. Our Hand-picked, AI optomized system has processed and summarized 27 articles from all over the internet to bring you the key 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. AI just passed a brutal finance exam most humans fail – should analysts be worried?
Frontier AI models are now able to pass a mock #CFA Level III exam, indicating they can apply advanced financial reasoning in #portfolio-management and #wealth-planning. The NYU Stern School of Business study analyzed 23 leading models from @OpenAI, @Google, @Anthropic, @Meta, @xAI, and @DeepSeek; OpenAI’s o4-mini led with 79.1%, followed by @Google’s Gemini 2.5 Flash at 77.3%, while the passing threshold is 63% and less than half (49%) of human test-takers passed in February. While most models scored around 71%-75% on the MCQ portion, their performance on the essay section varied more, highlighting that higher-order reasoning still differentiates frontier models from peers. The results suggest that simpler tasks have become commoditized, but complex analytical tasks requiring synthesis and professional judgment remain the area where AI outpaces most systems #professional-judgment. As AI capabilities advance, finance teams should monitor potential AI-assisted decision-making and implement guardrails to manage risks and opportunities.
2. Stars protest after AI actress gets agency interest: ‘What about living young women?’
An entirely virtual AI actress named Tilly Norwood, created by Xicoia and connected to Particle6, has attracted agency interest and intensified debate about rights and consent for living performers in AI-driven entertainment. Deadline reported agency interest, and reactions on social media from stars such as @MelissaBarrera, @MaraWilson, @NicholasAlexanderChavez, @LukasGage, and @LucyHale ranged from condemnation to sarcasm, with calls to consider the implications for living performers who could be composited or replaced. The discourse exposes a collision between #AI innovations in film and concerns expressed by #SAG-AFTRA and others about ownership of scans and the impact on real actors, echoing debates highlighted during the 2023 strike. Van der Velden, Particle6 founder, said the company plans to announce which agency will represent Norwood in coming months, underscoring that this is positioned as a test case for how talent representation might evolve with AI.
3. Larry Ellison on AI surveillance and Oracle’s TikTok deal
Larry Ellison, co-founder of Oracle, discusses Oracle’s role in handling #TikTok data surveillance and the broader implications of #AI in social media. He emphasizes Oracle’s capabilities in secure data management and fights concerns about government overreach in surveillance. Ellison argues that AI presents both opportunities and risks in content moderation and user data protection, highlighting the need for responsible technology use. This perspective links Oracle’s strategic positioning in the TikTok deal with evolving regulatory and technological landscapes in social media platforms.
4. Renewables vs SMR for datacenter energy: a debate on environmental and efficiency trade-offs
The debate between using #renewable energy and #SmallModularReactors (SMRs) for datacenters centers on balancing environmental impact and energy efficiency. Advocates of renewables emphasize the reduced carbon footprint and sustainability of solar and wind power, while proponents of SMRs highlight the consistent, high-density energy output and lower land usage. @Environmental experts caution about the intermittency of renewables requiring backup solutions, whereas nuclear energy, including SMRs, presents challenges like radioactive waste and higher upfront costs. The article discusses how integrating both approaches might optimize energy reliability and environmental responsibility for large-scale datacenter operations. Ultimately, the choice reflects varying priorities on climate goals, technological feasibility, and economic considerations.
@Jensen Huang says China is nanoseconds behind the US in chipmaking and argues Washington should stop walling off the market, contending that letting U.S. firms sell into China would serve American interests by spreading technology and extending geopolitical influence. In the BG2 podcast, he notes Nvidia’s bid to ship its H20 AI GPU to Chinese customers after a pause tied to export rules, with licenses reportedly being issued in August and a successor chip in development to comply with current restrictions. The piece highlights China’s push toward #selfSufficiency with Huawei’s Atlas 900 A3 systems and domestic silicon from #Baidu #Alibaba #Tencent #ByteDance that are CUDA-free and optimized for homegrown software, pressuring Nvidia’s dominance which once sat around 95% in China. Huang also frames China as publicly seeking an open market where international competition can thrive, while Nvidia seeks to maintain a foothold by staying engaged in China despite the geopolitical divide. The situation shows Nvidia navigating export controls while rival Chinese players accelerate their own chip programs.
6. Yamaha patents tech to replicate internal combustion engine sound and vibration in EV motorcycles
Yamaha has patented a system designed to replicate the sound and vibration of internal combustion engines in electric motorcycles, addressing common criticisms of electric vehicle riding experiences. The technology combines audio and haptic feedback to simulate engine noises and vibrations that traditional motorcycle riders associate with engagement and excitement. This system aims to bridge the sensory gap between gas-powered and electric motorcycles, potentially increasing acceptance among enthusiasts accustomed to combustion engines. As electric motorcycles gain popularity, Yamaha’s innovation could enhance rider experience by preserving the emotional connection to the machine. The development fits within broader trends in the motorcycle industry seeking to make electric models more appealing without sacrificing the classic feel.
7. Volvo Unveils BZr Electric Coach Chassis With New Battery And 480-kW Power Spec
Volvo has introduced the BZr electric coach chassis designed for US and Canadian markets, featuring a new 480-kW power output and a lithium-iron phosphate (LFP) battery pack developed with CATL, which is known for its safety and durability. The BZr is a full-size coach chassis equipped with an electric drivetrain capable of supporting various bodybuilders and operators, aiming to meet regional and intercity transit needs while achieving up to 340 miles of range on a single charge. Volvo’s partnership with CATL enables the use of LFP batteries, which provide enhanced safety compared to nickel-based alternatives commonly used in electric vehicles. This launch reflects Volvo’s commitment to expanding sustainable transportation options in North America, leveraging #electricvehicle technology and advanced #batterychemistry to reduce emissions and operating costs in coach services. The BZr chassis aligns with growing demand for #electricbuses and supports Volvo’s strategic goal to accelerate the adoption of zero-emission public transportation.
8. AI is helping judges to quickly close cases, and lawyers to quickly open them
Brazil is turning to #AI across its judiciary to handle a massive backlog, with judges closing cases faster and lawyers filing more with AI assistance. Since 2019, Brazil has developed or implemented over 140 AI projects that find precedents, categorize cases, and draft documents, and the Supreme Court department handles about 76 new cases per month amid a system of about 76 million lawsuits and $30 billion in annual costs (1.6% of GDP). A tool named MarIA, built by STF’s tech team and leveraging #Gemini and #ChatGPT, helps law clerks like @Arianne_Vasconcelos draft reports for judges such as @Luís_Roberto_Barroso. Experts warn that the drive could oversimplify the law to be computable, potentially undermining justice and equity, while the ‘vicious circle’ of AI increasing litigation persists, as noted by @Rodrigo Badaró. The article highlights a tension between efficiency gains and accountability as Brazil deploys one of the world’s largest AI programs in government #NationalCouncilOfJustice.
9. Many employees are using AI to create ‘workslop’
US workers are seeing AI-generated material enter their workflows, a phenomenon described as ‘workslop’ in a study by @Stanford and @BetterUp that finds 40% of US workers encountered such content in the last month. Workslop resembles impressive output but often lacks actionable facts, forcing colleagues to sort through it and convert it into usable work, costing about $186 per employee per month in lost productivity #AI #generativeAI #productivity. The impact is negative: more than half of recipients feel annoyed, over a third are confused, and nearly a quarter are offended, with 42% reporting less trust in the sender. The problem flows in both directions: 18% of workslop goes to managers and 16% originates from managers, showing a bidirectional pattern in corporate communication and the tech sector and professional services being major generators #workslop. Beyond individual cases, the article notes broader ROI questions, citing a UK government study that found no clear productivity boost from Copilot and MIT research suggesting about 95% of organizations see no measurable ROI from generative AI, underscoring that adoption does not guarantee gains.
Computer science students face challenges in job hunting despite the growth of #AI technologies, a situation that is often misunderstood as a direct consequence of AI hype. @Hany Farid, an expert in digital forensics and AI, explains that the issue stems more from a mismatch between students’ skills and job market demands rather than the rise of AI itself. Employers seek candidates with practical, relevant experience and applied problem-solving abilities, which many recent graduates lack. This disconnect suggests that students need to focus on gaining hands-on experience and adapting to industry needs. The situation highlights the importance of aligning education with evolving job market requirements rather than attributing job search difficulties solely to AI advancements.
11. And The Truth? This Writing Style Screams AI
The piece argues that AI writing is defined by a clipped, dramatic cadence that feels performative rather than conversational, making the style a clear AI fingerprint. It notes that six months ago AI could be spotted by polished grammar and emoji bullets, but the real giveaway is the rhythm and the one line paragraphs that split text like an infomercial. A Brief History of the AI Cadence traces it to speechwriters, preachers, and copywriters before GPT, with figures like @Reagan, @Clinton, @Obama, @Churchill, and @Blair, as well as @CaptainKirk, all using punchy rhythm to amplify emotion. The article explains that large language models are trained on vast amounts of spoken word material — transcripts of speeches, debates, and podcasts — rather than curated prose, which biases them toward this rhythm. Transcripts preserve spoken cadence and dominate training data, so models overindex this rhythm even when asked for varied prose, leading AI texts to read like motivational talks rather than journalism #AIcadence #LLMs #spokenword
14. AI can forecast your future health – just like the weather
AI can forecast future health risks decades ahead by spotting patterns in medical records, estimating the likelihood of more than 1,000 diseases in a weather-like forecast for health #AI #Delphi-2M. The model, named Delphi-2M, was trained on anonymous UK Biobank data including hospital admissions, GP records and lifestyle factors from over 400,000 people and later tested on Danish data from 1.9 million records, with predictions aligning with observed outcomes #UKBiobank #Denmark. It performs best for diseases with clear progression such as type 2 diabetes, heart attacks and sepsis, offering potential to identify high-risk patients for early intervention and to inform screening and resource planning #healthcare #diseaseprogression. Although not yet ready for clinical use and with potential biases from UK Biobank data, researchers including Prof @EwanBirney and Prof @MoritzGerstung see this as a new way to understand disease and move toward personalised care and scalable healthcare planning #personalisedcare.
15. High-Power Microwave System Downs 49 Drones in One Shot
A new high-power microwave (HPM) system has been developed to counter drone swarms by disabling 49 drones simultaneously through electromagnetic interference. The system emits a focused burst of microwaves that disrupts the electronics of multiple drones at once, rendering them inoperative without physical damage to drones or surrounding infrastructure. This technology offers a scalable and efficient approach to neutralizing threats posed by drone swarms in military and security contexts, providing an alternative to traditional kinetic or explosive methods. The development marks a significant advancement in #electronicwarfare, showcasing potential for controlling aerial threats with less collateral damage. As drone usage increases across sectors, such #counterdrone solutions become crucial for maintaining operational safety and defense.
The AI surge mirrors the late-1990s dot-com boom, with valuations and investments rising on transformative potential rather than current profitability. Global corporate AI investment reached $252.3 billion in 2024, and America’s tech giants plan $320 billion in AI infrastructure spending this year, trends that have even @Sam Altman and @Mark Zuckerberg acknowledging the parallels. Yet the dot-com crash was triggered by a combination of higher interest rates and a global recession, exposing many internet firms with fundamentally flawed business models like Commerce One, TheGlobe.com, and Pets.com. The best-known parallel is infrastructure overbuild: massive fiber optic deployment driven by inflated growth claims, which amplified the downturn when demand didn’t materialize. The takeaway is to demand sustainable profitability and sound business models for #AI, using lessons from history to separate hype from durable value.
17. Tech Leaders Warn of AI-Induced Jobs Crisis
Top figures in AI, including @DarioAmodei of #Anthropic and @JakeVandeHei, express concern that rapid advances in artificial intelligence could trigger a global jobs crisis. They point to automation accelerating beyond prior expectations, potentially displacing millions in sectors like manufacturing and services. The dialogue highlights the urgent need for policymakers to develop strategies addressing workforce transitions and social safety nets. This caution underscores the broader challenge of integrating AI while safeguarding economic stability. These perspectives call for balanced approaches combining technological progress with proactive labor market interventions.
18. How South Korea plans to best OpenAI, Google & others with homegrown AI
South Korea aims to compete with global AI leaders like @OpenAI and @Google by investing heavily in developing homegrown AI technologies tailored to local needs. The government is promoting innovation through substantial funding, fostering partnerships between academia, industry, and government entities. This strategy includes prioritizing data privacy, ethical AI use, and cultivating domestic talent to strengthen national competitiveness in AI development. By leveraging these focused initiatives, South Korea intends to create unique AI solutions that align with its cultural and economic priorities. This approach positions the country to carve out a distinctive role in the global AI landscape while supporting sustainable technological growth.
19. A new buzzword is hanging over businesses as they rush into AI
As businesses rapidly adopt #AI technologies, a new term, ‘AI ethics,’ has emerged to address concerns around responsible AI use. Companies are grappling with challenges related to data privacy, bias, and transparency in AI systems, highlighting the need for ethical frameworks. Industry leaders like @SatyaNadella emphasize balancing innovation with responsibility to maintain trust and ensure equitable AI applications. The growing focus on AI ethics reflects a broader movement to regulate and standardize AI practices, preventing misuse and fostering sustainable growth. This buzzword encapsulates the critical intersection of technology advancement and moral accountability in the AI era.
20. Famed Roboticist Says Humanoid Robot Bubble is Doomed to Burst
Humanoid robots have garnered significant hype, but the enthusiastic expectations surrounding them are unsustainable according to a renowned roboticist. This expert highlights that while advancements in #robotics have progressed, the complex challenges of creating functional humanoid machines remain vastly underestimated. Evidence includes the technical difficulties in replicating human dexterity, perception, and decision-making in machines which have limited commercial viability so far. The analysis points to a speculative bubble fueled by media excitement and investment hype, potentially leading to a market correction. The expert’s perspective urges a more cautious and realistic approach to humanoid robotics development and investment.
21. Apple is reportedly testing an AI chatbot for Siri called Veritas
Apple is developing an AI chatbot named Veritas to enhance its Siri virtual assistant, aiming to provide more accurate and conversational responses. According to reports, this effort involves integrating advanced conversational AI capabilities beyond Siri’s current technology, potentially improving user interaction by offering detailed and context-aware answers. The initiative signals Apple’s push to compete with other AI assistants like ChatGPT and Google Bard, focusing on privacy and seamless integration within its ecosystem. Veritas could change how users interact with Apple devices by making Siri more intelligent and useful in handling complex queries. This development aligns with industry trends where major tech companies leverage #AI to redefine digital assistants and user experience.
22. Apple’s iPhone 17 will forever change how we take selfies – including on Android phones
Apple’s iPhone 17 redefines selfies by upgrading the front-facing camera across the lineup, moving from a 12MP rectangular sensor to a 24MP square sensor that can produce an 18MP image in either horizontal or vertical orientation. A new #CenterStage feature appears when you open the front camera, letting users switch between vertical and horizontal framing without flipping the device, and you can enable #AutoZoom and #AutoRotate in the settings. Apple notes the front camera has long been a weak point while also illustrating how central selfies are to usage, citing hundreds of billions of selfies and a large installed base of iPhone users. All four iPhone 17 models receive the same upgrade, signaling a stronger emphasis on front-cam quality. The changes are positioned to set a standard that other manufacturers may imitate in 2026, influencing how #selfie cameras evolve across Android and iPhone alike, as seen in public moments with @Tim Cook at the 2025 Apple Event.
23. YouTube (GOOGL) Tests AI Music Hosts For Premium Users, Taking Aim At Spotify (SPOT)
YouTube is experimenting with AI-generated music hosts for its Premium subscribers, aiming to enhance the user experience and challenge competitors like Spotify. This feature uses artificial intelligence to create personalized DJ-style hosts who introduce songs and add commentary, increasing engagement on the platform. The implementation reflects a broader trend in streaming services leveraging #AI and #machinelearning to differentiate offerings and attract premium subscribers. By integrating AI hosts, YouTube seeks to combine music streaming with interactive elements, potentially boosting user retention and subscription growth. This initiative highlights the ongoing competition in the music streaming market to innovate and expand audience appeal.
24. MSI Confirms MAX and EVO Refresh Motherboards Feature 64MB BIOS
MSI has confirmed that its MAX and EVO refresh motherboards will come with a 64MB BIOS, a significant upgrade over previous versions. This increase in BIOS size allows for enhanced support for upcoming CPUs and additional features, ensuring better compatibility and functionality. By expanding the BIOS capacity, MSI addresses limitations that affected older motherboards, improving user experience and future-proofing the platforms. This move aligns with broader industry trends where motherboard manufacturers enhance BIOS storage to support expanding hardware requirements. Consequently, MSI’s decision reflects a commitment to providing users with reliable, updated platforms suited to evolving PC components.
25. Lufthansa to Cut 4,000 Administrative Jobs by 2030 Amid AI Push
Lufthansa announced plans to reduce 4,000 administrative jobs by 2030 as part of a strategic move to embrace #artificialintelligence and improve efficiency. The airline expects AI technologies to automate a significant portion of routine administrative tasks, allowing it to streamline operations and reduce costs. This shift reflects a broader trend in the aviation industry, where companies leverage AI for competitive advantage and operational optimization. By prioritizing digital transformation, Lufthansa aims to better position itself for future market demands while managing labor expenses. The decision underscores the growing impact of AI on workforce structures in major corporations.
26. As AI threatens white-collar work, more young Americans choose blue-collar careers
Young Americans are increasingly pursuing blue-collar and skilled-trades careers as AI begins to encroach on entry-level white-collar work and the rising cost of college debt influences career choices. In one example, @JacobPalmer, 23, started Palmer Electrical in 2024 after an apprenticeship and is projected to generate more than $150,000 in revenue this year, while a ResumeBuilder.com survey finds that 42% of Gen Zers are in or pursuing blue-collar jobs. Industry voices say the stigma around trades is fading, with @DavidAsay noting that a cool, in-demand skillset now has appeal and @AngieHicks highlighting the changing perception of trades. The cost of college remains burdensome, with average annual costs over $38,000 and private costs near $60,000, and the lifetime cost of a bachelor’s degree can exceed $500,000; meanwhile, unemployment for 23-to-27-year-old grads is around 4.6% this year, up from 3.2% in 2019, while non-college-educated peers have seen a much smaller uptick, by roughly 0.5 percentage points. In solar work, firms like @VinnieCurcie’s OC Solar suggest AI may soon handle much of sales and project management, even as on-site field work remains essential. As AI expands, trades offer independence and clearer pathways, potentially reshaping the labor market toward #blue-collar careers and other skilled trades.
27. Reddit Seeks to Strike Next AI Content Pact With Google, OpenAI
Reddit is negotiating new agreements with @Google and @OpenAI to secure compensation for the use of its user-generated content in AI training. These discussions follow prior deals Reddit struck to monetize access to its vast dataset as #AI developers increasingly rely on real-world content to improve their models. The company aims to set industry precedent on content licensing, ensuring creators and platforms are fairly paid amid the surge in generative AI development. By establishing clearer terms, Reddit hopes to balance user trust with innovation, illustrating the ongoing shift in how digital content fuels AI advancements. This move highlights broader trends where content platforms assert intellectual property rights in the rapidly evolving #AI ecosystem.
28. LockBit’s new variant is ‘most dangerous yet’
LockBit 5.0 is the most dangerous yet, as it targets #Windows, #Linux, and #VMware_ESXi in a single campaign, amplifying enterprise-wide impact across endpoints, servers, and virtualization hosts. Trend Micro’s technical breakdown shows heavy obfuscation and improved evasion across variants, with #Windows payloads loaded via DLL reflection and anti-analysis packing, #Linux variants using CLI directives to tailor targets, and ESXi builds designed to seize virtualization infrastructure by encrypting VMs. Each encrypted file receives a random 16-character extension to complicate restoration, and the modular architecture plus multi-OS encryption routines enable attackers to shorten the breach-to-encryption window across the enterprise. The revival follows a law-enforcement takedown (Operation Cronos), but the group is reactivating its affiliate network under a rebranded, hardened platform, signaling resilience and cross-platform expansion. Defenders face elevated risk as LockBit 5.0 can terminate security processes and delete backups, with ESXi-focused attacks threatening recovery, so comprehensive cross-platform defenses that protect virtualization infrastructure are essential to countering a campaign spanning #Windows, #Linux, and #VMware_ESXi.
29. ‘You’ll never need to work again’: Criminals offer reporter money to hack BBC
A BBC cyber correspondent @Joe Tidy reveals how criminals tried to recruit insiders to hack the BBC, offering 15% of any ransom in exchange for access to his PC. The group, calling itself Medusa and led by a ‘reach out manager’ named Syn, claimed they could extort tens of millions and offered a 25% share of the final negotiation as they siphoned 1% of the BBC’s revenue, while promising to delete the chat to avoid discovery. Tidy points to insider risk and cites prior cases where IT workers sold login details and recent hacks of a UK healthcare company and a US emergency services provider as evidence of these schemes. Medusa operates as #ransomware-as-a-service, with administrators believed to be in Russia or nearby states and to target mainly non-Russian entities, while the article notes the National Crime Agency guidance not to pay and calls for stronger monitoring of insider access.
That’s all for today’s digest for 2025/09/29! We picked, and processed 27 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! 🚀