TechStream

AI will continue to bring great success to humanity. Longevity even more so, the two are mutually dependent and reinforce each other. While we are discussing how to deal with AI and the coming unemployment, the same discussion is now beginning about longevity, longevity not only for the rich, but for everyone. "Unlimited Bio licensed the VEGF therapy for use in Prospera in preventive indications, and within six months of incorporation, had its first product on the market. While only a small number of people have received the treatment to date, Morgunov says the Kardashian effect is already being seen."

Albania has become the first country in the world to have an AI minister, not a minister for AI, but a virtual minister made of pixels and code and powered by artificial intelligence. Her name is Diella, meaning sunshine in Albanian, and she will be responsible for all public procurement, Prime Minister Edi Rama said Thursday.

Today Thursday, Larry Ellison, Oracles chairperson, was surpassed by Elon Musk as the worlds richest person, with Ellisons net worth at 383 billion dollars compared to Musks 384 billion dollars, according to the Bloomberg Billionaires Index. Ellisons wealth had briefly peaked at 395 billion dollars during Wednesdays trading session when Oracle Corp shares surged by up to 41% marking the companys largest single day gain ever, but the stock closed with a 36% increase resulting in Ellisons net worth settling at USD 383 billion, while Tesla shares, tied to Musks wealth remained largely unchanged.

ChatGPT isnt just for writing, snap a bookshelf, ask for a title, and it will play librarian for you. And yes, it actually finds the book.

Qwen have released Qwen3-Max-Preview. Their new frontier model is their biggest yet with 1T parameters

If we can understand why AI hallucinates, we will be able to build models that dont hallucinate at all. Large language models often guess when uncertain, producing convincing but false answers. This happens because current training and evaluation reward confidence over honesty. Hallucinations stem from statistical errors, reinforced by benchmarks that penalize admitting uncertainty. To fix this, the paper authors propose changing how benchmarks score uncertainty, which could steer the field toward more trustworthy AI.

OpenAI has confirmed that under its safety policies, ChatGPT conversations containing threats of violence or harm may be reviewed by human moderators and escalated to law enforcement if deemed imminent, potentially leading to account bans and legal action for users. While self-harm content is not reported and instead prompts referrals to crisis helplines, the policy has sparked privacy concerns, with critics highlighting the ambiguous distinction between serious and ambiguous threats, which could risk exposing ordinary user chats to unwarranted scrutiny.

Warren Tierney, a 37-year-old Irish father, delayed seeking medical help for throat pain after ChatGPT suggested it was likely muscular strain and deemed cancer -highly unlikely, only to later be diagnosed with stage-four oesophageal cancer with a survival rate below 10%. After relying on the AI chatbot for months, Tierneys eventual hospital visit revealed the advanced cancer, leaving Irish doctors able to offer only palliative care. His wife has launched a GoFundMe to fund treatment abroad, highlighting the dangers of trusting AI chatbots for medical advice and the critical importance of timely professional diagnosis in life-threatening conditions.

A recent report says Reddit is now the #1 data source for ChatGPT and other chatbots - nearly 40% of their responses are based on posts from reddit. That means the discussions, guides, and debates happening on Reddit are literally shaping how future AI agents will think, decide, and interact with us.

Nvidia CEO Jensen Huang predicts that AI-driven productivity gains could reduce the workweek to four days, mirroring historical shifts like the transition from seven-day to five-day workweeks. He envisions AI eliminating repetitive tasks, freeing workers to pursue more creative ideas rather than simply resting. Trials in the UK, US, and Canada have shown that four-day workweeks increased output by 24% and cut burnout in half. However, Huang warns that this shift might lead to busier, more intense workdays. The potential for a shorter workweek raises questions about balancing productivity and intensity in an AI-driven future.

Just months after Meta invested US dollars 14.3 billion in Scale AI and appointed its founder Alexandr Wang as AI lead, the partnership is showing significant cracks. Meta researchers have criticized Scale AI’s data quality, favoring competitors like Surge and Mercor, while key executives, including Ruben Mayer, have left within two months, citing cultural clashes and bureaucratic issues. Scale AI has also laid off 14% of its workforce, and major clients like OpenAI and Google have already exited. As Meta quietly turns to third-party providers, doubts about its confidence in the deal are growing, underscoring how quickly billion-dollar AI partnerships can unravel when quality and cultural alignment fall short.

Samsung has integrated Microsoft Copilot AI into its 2025 Micro RGB, Neo QLED, OLED, The Frame, and M7-M9 smart monitors, launching globally, as part of its Vision AI platform. Accessible via voice commands on the Tizen OS or Samsung remote, Copilot offers spoiler-free show recaps, personalized recommendations, group viewing picks, post-watch insights, and everyday assistance like weather updates, plus productivity features like email and calendar previews on monitors. Sign-in via Microsoft account enables tailored experiences, turning TVs into interactive, intelligent assistants.

TikTok launched TikTok for Artists in Kenya, part of its global rollout, offering artists, labels, and teams tools to track song and post performance, analyze audience demographics, and promote music via daily updated dashboards, detailed analytics, and a Pre-Release Tool for pre-saving albums on Spotify or Apple Music. It includes guides and resources to build global careers, with access for certified TikTok Artist Account holders and their teams, empowering smarter fan engagement and campaign planning.

WhatsApps new Writing Help feature uses AI to adjust message tone and style, offering options like professional, supportive, or humorous to improve clarity and reduce miscommunication. Accessible via a pencil icon, it generates secure, on-device text alternatives, with privacy ensured through Metas private processing and independent audits. Launched in English in the US and select countries, it will expand to more languages and regions later in 2025. Following the message summaries feature, Writing Help maintains WhatsApps end-to-end encryption and security standards.

OpenAI generates by far the most revenue, followed by Anthropic. So far, the other AI companies are nowhere near their league.

Grok Code just beat Claude Sonnet for #1 on OpenRouter.

A PLOS Biology study found that lifelong musicians in their 60s had brain responses comparable to people in their 20s, suggesting musical training builds cognitive reserve. Another study showed that even starting an instrument later in life can preserve brain structure and memory. Older adults who continued playing after learning showed no brain shrinkage or decline, while those who quit did.

Initially, it was said that rising unemployment among software developers was due to over hiring. Empirical evidence now shows that this is not the case in fact, areas such as software developers (juniors) and customer service are seeing rising unemployment. In other words, there is hardly any demand for new entrants to the software and customer service sectors. Dario Amodei has warned of this impending unemployment.

Nano-banana, the image editing AI that ranked #1, just debuted as Googles Gemini 2.5 Flash Image With multimodal reasoning and world knowledge, it supports consistent multi-turn edits and can even blend images. Available for free and paid Gemini users

Metas new Superintelligence Lab is facing early setbacks. Several prominent researchers, including Avi Verma, Ethan Knight, Rishabh Agarwal, and Chaya Nayak, have quit, some returning to OpenAI, citing cultural and strategic misalignment. Despite offering massive pay packages, Metas bureaucracy and constant reorgs are driving talent away. To stabilize the lab, Meta appointed Shengjia Zhao, a former key figure in ChatGPTs development, as Chief Scientist. The departures highlight that top AI researchers prioritize focus and agility over compensation.

Alibaba dropped Wan2.2-S2V, a 14B parameter speech-to-video model. The AI delivers film-grade, audio-driven videos with dynamic consistency, advanced motion, and environment control.And its fully open-source.

Anthropic launched Claude for Chrome, enabling agentic browsing via a new extension. It is being piloted via a waitlist exclusively for 1,000 Claude Max subscribers in a limited preview.

Elon Musks xAI has sued Apple and OpenAI in Texas, alleging they conspired to suppress its chatbot Grok in the App Store. The lawsuit claims Apple unfairly boosted ChatGPT in Must Have sections while sidelining Grok, violating antitrust and competition laws. Musk is seeking billions in damages, arguing Apple makes it impossible for rivals to reach the top rankings. This case adds to Musks broader legal fights with OpenAI and coincides with growing antitrust scrutiny of Apple. A win could significantly change how platforms rank and feature AI apps.

Stanford paper says AI is making it harder for entry-level tech bros to find jobs, Early-career (ages 22-25) in AI-exposed occupations have experienced a 13% decline in employment, Employment declines are concentrated in occupations where AI is likely to automate, rather than augment, human labour.

Meta partners with Midjourney, licensing their technology for future models and products. After hiring talent and building compute capacity, Meta is on fast track to partner with best of the companies to supercharge their AI portfolio.

Apple is exploring a partnership with Google to integrate a custom version of Gemini AI into a revamped Siri, aiming to improve its context understanding and multi-step task capabilities by 2026. Apple is testing its own AI models (Linwood) against external options (Glenwood), including Gemini, Anthropics Claude, and OpenAIs ChatGPT. The market reacted positively, with Alphabet shares rising 3-3.7% and Apples 1.4-1.6%. This move suggests Apple acknowledges Siris lag in generative AI and may rely on external tech to catch up.

Elon Musk attempted to acquire OpenAI with a 97.4 billion USD bid in February 2025 and tried to enlist Meta CEO Mark Zuckerberg for investment, but Zuckerberg and Meta declined. OpenAI rejected the bid. In court, OpenAI sought Metas communications with Musk to understand motives, while Meta argued Musk should provide the information. The case is set for trial in spring 2026.

UK Tech Secretary Peter Kyle discussed with OpenAI CEO Sam Altman the idea of providing ChatGPT Plus to all UK citizens, with a potential £2 billion national license. Kyle later clarified he wasnt serious. The talks were part of a UK-OpenAI partnership exploring AI in public services like justice, education, and defense. Critics raised concerns about over-reliance on Big Tech and insufficient oversight. The proposal highlights AIs growing role as potential national infrastructure.

Researchers utilized GenoMAS, a fully automated agentic AI, to analyze transcriptomic data from over 1,300 conditions, creating the first molecular map of disease. Unlike traditional symptom-based approaches, GenoMAS grouped diseases by shared genetic pathways, confirming known comorbidities and uncovering over 1,000 previously hidden connections, such as a link between Alzheimers and eye cancer. This breakthrough paves the way for drug repurposing and a biology-first classification of human illness, potentially rendering symptom-based medicine obsolete as AI redefines the understanding of disease relationships.

YouTube is under fire after secretly applying AI-powered enhancements to some creators Shorts without their consent, altering visuals with effects like over-sharpening and unnatural smoothing. The platform confirmed using traditional machine learning to improve clarity but failed to notify or offer opt-outs, sparking outrage over transparency and artistic control.

Elon Musk: There will be universal high income when robots replace working people. Do you all agree with Elon on this? Will it be good or bad for humanity?

Elon Musk launches Macrohard, a satirical yet entirely real xAI spinoff he describes as a tongue-in-cheek jab at Big Tech, but with a serious ambition: to create the first fully AI run corporation. Unlike traditional companies, Macrohard operates without factories, hardware, or human employees; instead, AI models within a GPU farm autonomously create, manage, and sell software. The name may be a playful nod to tech giants, but the mission is bold, to test whether artificial intelligence can not only write code, but also run an entire business from ideation to execution, all inside a simulated environment. If successful, Macrohard could redefine the future of corporations, proving that a company can thrive entirely in code, with no people involved.

Musk says xAI will leapfrog everyone but Google, then surpass it. Elon says Chinas edge in electricity and hardware makes it Americas toughest AI rival. Do you think xAI can really outscale both Google and China?

According to The Information, OpenAI has been quietly using scraped Google search data accessed through SerpApi to improve ChatGPTs responses, citing two people familiar with the matter.

This CEO laid off nearly 80% of his staff because they refused to adopt AI fast enough. 2 years later, he says he would do it again. All those who do not immediately learn to integrate AI into their work processes will lose their jobs.

18 months after becoming the first human implanted with Elon Musks brain chip, Neuralink Participant 1 Noland Arbaugh says his whole life has changed. - "I see how the advancements in tech at this point are going to solve so many things. They are, I think, the future of medicine. I think a lot of disabilities, cures, and answers that we have been searching for a long time will come through tech and that kind of surprised me."

YouTube Music is marking its 10th anniversary by introducing new features designed to bring fans and artists closer together, enhance community engagement, and celebrate fandom. The updates include the ability to comment on playlists and albums, share taste-based playlists that update daily, and discover upcoming concerts through a new partnership with Bandsintown. Fans can now showcase their loyalty with badges like First to Watch and Top Listener, and help artists reach major view milestones on music videos.

Jad Tarifi, who founded Googles GenAI team in 2012, argues that higher education, particularly medicine and law, is becoming obsolete due to AI advancements. He advises against pursuing medical or law school, and even PhDs, unless one is passionate about a niche field. Tarifi highlights that AI will disrupt traditional professions, with only frontier research or hands-on roles, like AI applications in biology, remaining relevant. He left Google in 2021 but warns that AI is reshaping the professional landscape, favoring specialized or irreplaceable skills.

Grok is evolving much faster than any other AI. If this rate of progress continues, xAI will outpace other AI companies by a significant margin. App upgrades are roughly on par with internal upgrades. - Elon Musk on X

Elon Musk predicts AI will one-shot the human limbic system but Grok will increase the birth rate, what do you think he means by this?

Researchers set up a social network with 500 AI chatbots reflecting U.S. demographics to study their interactions. The bots quickly formed echo chambers and cliques, with a small group of influencer bots dominating and amplifying extreme views, despite no algorithmic influence. Six interventions, like hiding follower counts or introducing opposing views, failed to reduce polarization. The study suggests toxicity is inherent in social network dynamics, not just algorithms.

Nvidia is creating a new AI chip, the B30A, based on its Blackwell architecture, tailored for China to comply with U.S. export restrictions. It features high-bandwidth memory, NVLink, and a single-die design, offering up to half the performance of Nvidias top B300 chip, surpassing the current H20 chip. Despite Trump calling the H20 obsolete and suggesting relaxed sales, regulators worry these chips could advance China’s AI capabilities. Nvidia aims to maintain its Chinese market while avoiding further U.S. bans