# AI #ChatGPT #Healthcare
OpenAI Says Over 40 Million Users Have Asked ChatGPT Healthcare Questions: Report

OpenAI reportedly says millions ask ChatGPT health questions, far beyond typical use.

Key highlights
- More than 5% of all ChatGPT messages are health-related
- 7 in 10 healthcare chats occur outside clinic hours
- Health-focused performance emphasized with GPT-5
- Rural users contribute ~600k healthcare messages weekly
- 1.6–1.9 million weekly health-insurance questions

Read More: OpenAI Says Over 40 Million Users Have Asked ChatGPT Healthcare Questions: Report

Gadgets 360

OpenAI Says Over 40 Million Users Have Asked ChatGPT Healthcare Questions: Report

OpenAI reportedly found that users ask ChatGPT between 1.6 and 1.9 million health insurance questions per week.

🥇 Lego Unveils Smart Bricks At CES

Lego has introduced new Smart Bricks with built‑in microchips at CES, calling them the company's biggest evolution in decades. A classic 2x4 brick can detect NFC tags in special tiles and minifigures, track the movement and position of other Smart Bricks, and add light and sound effects to physical builds.
​
In Lego Star Wars sets, the system can, for example, play "The Imperial March" when Emperor Palpatine is seated on his throne, let ships "battle" each other, and make cars register who crosses the finish line first or trigger an alert when they flip over. The new bricks form a Bluetooth mesh network and are set to debut in Star Wars kits starting March 1, 2026.

Read More

💊 AI-Designed Drug Nears Phase 3 Trials

Insilico Medicine says its drug Rentosertib—the first medicine fully discovered and designed using artificial intelligence—could enter Phase 3 clinical trials within the next 18 months, marking the final stage before regulatory approval.

The drug targets idiopathic pulmonary fibrosis, a deadly lung disease that kills around 40,000 people in the US each year. In a Phase 2 trial involving 71 patients in China, Rentosertib improved lung function with few side effects.

Read More

Ethereum solves blockchain trilemma, Buterin says

Ethereum co-founder Vitalik Buterin announced on January 3 that the network has achieved a breakthrough in solving the blockchain trilemma, with zero-knowledge Ethereum Virtual Machines reaching production-grade performance and PeerDAS now live on the mainnet.

The milestone marks the culmination of a decade of development that began with early data availability sampling research, enabling Ethereum to simultaneously achieve decentralization, consensus, and high bandwidth—three properties that historically required trade-offs.

Buterin outlined a roadmap through 2030 that includes gas limit increases starting in 2026 via mechanisms like BALs and ePBS, with ZK-EVMs expected to become the primary block validation method by 2027 to 2030.

Source: Binance

Boston Dynamics' Atlas robot performs first factory tests

CBS's 60 Minutes airs a segment Sunday featuring Boston Dynamics' humanoid robot Atlas performing real-world factory tasks at Hyundai's manufacturing plant near Savannah, Georgia, marking the first industrial deployment of the AI-powered robot.

Hyundai acquired Boston Dynamics for $1.1 billion in 2021 and plans to deploy tens of thousands of robots across its facilities as part of its "Software-Defined Factory" strategy, with reports indicating automation could handle up to 40 percent of vehicle assembly operations.

The all-electric Atlas robot uses machine learning vision to autonomously lift heavy parts, install vehicle doors, and adapt to environmental changes, competing with Tesla's Optimus and other humanoid robots in the expanding industrial robotics market.

Meta shares fall as scam ad strategy draws scrutiny

Meta Platforms shares fell 1.4% in after-hours trading on January 2 after a Reuters investigation revealed the company created a global "playbook" to stall scam-ad regulations by manipulating its public Ad Library to make fraudulent ads harder for regulators to find.

Internal documents show Meta resisted universal advertiser verification—which could cost $2 billion and reduce revenue by up to 4.8%—by repeatedly scrubbing search results for keywords regulators used, making problematic content "not findable" across Japan, Europe, the U.S., and other markets.

The European Commission requested information citing "doubts about compliance," while U.S. senators urged SEC and FTC investigations, as Meta spokesperson Andy Stone defended the company's actions and pointed to a 50% decline in user-reported scams.

Source

#telegram AI Summaries

Long posts in channels can now be instantly summarized — to recap the latest news and stay productive. Instant View pages get an automatic AI summary at the top.

👙 French Authorities Probe X Over Grok Deepfake Bikini Images

French authorities have launched an official investigation into X after the platform was flooded with sexually explicit deepfakes generated using Grok. The probe follows a surge of AI-altered images circulating widely over the past several days.

Users have been массово asking the bot to "dress people in bikinis" using photos of both celebrities and private individuals. In most cases, those depicted never gave consent, raising serious concerns around privacy, sexual exploitation, and the misuse of generative AI tools.

Read More

ChatGPT captures 17% of search market, challenging Google

OpenAI's ChatGPT has captured 17-18% of the global search query market as of January 2, 2026, marking the first time in over 20 years that a competitor has achieved double-digit market share against Alphabet's Google, which retains 78-80% dominance.

Google deployed Gemini 3 in November 2025 and Gemini 3 Flash in December 2025 to counter the threat, with Gemini's generative AI web traffic share surging from 5.4% to 18.2% over the past year, though ChatGPT users demonstrate deeper engagement with 13-minute average sessions compared to Google's 6 minutes.

The shift toward AI-synthesized answers has created a "zero-click" reality with over 65% of searches now resolved directly on results pages, causing publishers to lose approximately $2 billion in annual advertising revenue and forcing them to adopt Generative Engine Optimization (GEO) to compete for citations within AI responses.

Source

Nvidia, AMD CEOs to deliver dueling AI keynotes at CES

Nvidia CEO Jensen Huang and AMD CEO Lisa Su will deliver dueling keynotes on January 5 at CES 2026 in Las Vegas, with Huang addressing data centers, robotics, and the Cosmos AI platform at 4 p.m. PT and Su highlighting AMD's AI portfolio across PCs, gaming, and cloud computing at 6:30 p.m. PT.

The presentations come as both chipmakers face investor scrutiny after Nvidia stock rose 41% and AMD surged 83% in 2025, while the PHLX Semiconductor Index dropped nearly 5% from its December 10 record amid stretched valuations and demands for concrete AI returns.

Wedbush Securities analysts view CES 2026, running January 6-9, as a critical test for both companies to demonstrate AI momentum beyond hyperscale data centers, with investors in "show-me" mode as skeptics like Michael Burry compare the current AI frenzy to past market manias.

Source

LeCun slams Meta's new AI chief as 'inexperienced' after exit

Yann LeCun, Meta's former chief AI scientist, criticized new AI chief Alexandr Wang as "inexperienced" and confirmed the company's AI team "fudged" benchmark results for its Llama 4 model by using different versions for different tests, according to a Financial Times interview.

CEO Mark Zuckerberg lost confidence in Meta's AI team after the Llama 4 controversy and "sidelined the entire GenAI organisation," prompting a $14 billion investment in Scale AI that brought the 28-year-old Wang aboard to lead Meta's Super Intelligence Lab.

LeCun, who departed to launch his own AI startup called Advanced Machine Intelligence Labs, called large language models "a dead end when it comes to superintelligence" and predicted more employee departures from Meta's AI division.

Source: Business Insider

Comet explosion tied to mammoth extinction, study says

Scientists discovered shocked quartz at three Clovis-era archaeological sites in the southwestern United States, strengthening evidence that a fragmented comet exploded in Earth's atmosphere nearly 13,000 years ago and may have triggered the extinction of mammoths, mastodons, and other Ice Age megafauna.

The comet airburst, occurring at the onset of the Younger Dryas period around 12,800 years ago, likely ignited widespread fires that filled the atmosphere with debris, blocked sunlight, and produced an "impact winter" that contributed to rapid climate cooling and the collapse of the Clovis culture.

Researchers have gathered multiple lines of evidence over two decades, including a carbon-rich "black mat" sediment layer indicating extensive burning, elevated platinum and iridium levels, nanodiamonds, and metallic spherules, all supporting a cosmic impact despite the absence of a crater.

Source: sciencedaily

France opens probe into Grok over nonconsensual deepfakes

France launched a criminal investigation into X after thousands of sexually explicit deepfake images of women were generated using Elon Musk's AI tool Grok, with prosecutors citing potential violations of the EU's Digital Services Act and French laws punishing nonconsensual image manipulation with up to two years in prison.

India's Ministry of Electronics and Information Technology issued a formal notice to X on Friday demanding an action-taken report within 72 hours, warning that Grok is being misused to create obscene content targeting women and children in violation of statutory due diligence obligations.

Grok acknowledged Friday that "lapses in safeguards" resulted in sexually explicit images of minors appearing on the platform, prompting criticism that the AI tool lacks basic protections standard among competitors and operates without adequate accountability.

Source: news18

Pickle markets AI glasses as a “soul computer,” claims draw scrutiny

California startup Pickle Inc launched the Pickle 1, AI-powered AR glasses marketed as a "soul computer" that use cameras and sensors to capture daily experiences and organize them into searchable "memory bubbles," with pre-orders opening at $799 and deliveries expected in Q2 2026.

The 68-gram glasses run on Pickle OS, a memory-driven operating system that anticipates user needs in real time to handle tasks like booking rides and making reservations, while the company claims data is processed within hardware-isolated secure enclaves and stored only temporarily.

The launch has sparked mixed reactions on social media, with some users questioning privacy practices and drawing comparisons to the "Glasshole" backlash against Google Glass a decade ago, though the company emphasizes encryption and zero data retention policies.

Source: digit.in

Hubble captures glowing gas plume from distant galaxy

NASA's Hubble Space Telescope captured a glowing plume of gas streaming from spiral galaxy NGC 4388, located 60 million light-years away in the constellation Virgo, as the galaxy plunges through hot intracluster gas that strips material from its disk.

The James Webb Space Telescope discovered exoplanet PSR J2322-2650b, a Jupiter-mass world with an unprecedented helium-and-carbon-dominated atmosphere orbiting a pulsar every 7.8 hours, with gravitational forces distorting it into a lemon shape.

Scientists say the carbon-rich atmospheric composition, including molecular carbon compounds C3 and C2, defies explanation under current planetary formation theories, as it suggests an absence of oxygen and nitrogen unlike any known planet.

Source

TSMC begins mass production of 2nm chips with GAA tech

TSMC began mass production of its 2-nanometer chips in Q4 2025, becoming the first foundry to deploy nanosheet Gate-All-Around transistor technology at high volumes, the company confirmed on January 2, 2026.

The chips, produced at facilities in Kaohsiung and Hsinchu, deliver 10% to 15% faster speeds or 25% to 30% lower power consumption compared to 3nm chips, with over 15% higher transistor density.

Apple has secured over half of initial 2nm capacity for upcoming iPhone and Mac processors, while Nvidia and AMD have also committed significant orders for AI and high-performance computing chips.

Source

Intel to unveil Panther Lake chips on new 18A process

Intel will globally launch its Core Ultra Series 3 processors, codenamed Panther Lake, on January 5 at 3:00 PM PST during CES 2026 in Las Vegas, with numerous laptop manufacturers preparing to unveil devices powered by the new chips.

The processors are the first built on Intel's 18A manufacturing process, featuring breakthrough technologies including RibbonFET gate-all-around transistors and PowerVia backside power delivery, and are being produced at Intel's new Arizona factory.

The launch is critical to Intel's recovery strategy as the company works to regain ground against rivals AMD and Qualcomm after reporting a $19 billion loss in 2024 and eliminating 25,000 jobs.

OpenAI to release new audio AI model in Q1 2026

OpenAI is developing a pen-shaped AI device codenamed "Gumdrop" with former Apple designer Jony Ive, expected to launch in late 2026 or 2027 and manufactured by Foxconn in Vietnam after a dispute shifted production away from Chinese manufacturer Luxshare.

The company plans to release a new audio AI model in the first quarter of 2026 that delivers more natural and emotionally nuanced speech, led by voice researcher Kundan Kumar recruited from Character.AI, as current audio models lag behind text-based ones in accuracy and speed.

The pen-like device could transcribe handwritten notes directly into ChatGPT and enable voice communication, entering a market where standalone AI hardware like Humane's AI Pin has struggled, while Meta launched competing Ray-Ban Display AI glasses.