What deserves your attention.

Medium launched TK, a web writing app that integrates an AI assistant into the editor. Users can call the tool by typing “@TK” and ask for help answering questions, retrieving saved passages, organizing ideas and suggesting prompts. Medium says TK also offers editorial and headline feedback, as well as tools to match drafts to the style guide of publications on the platform.

The launch comes amid debate over AI use in literature and journalism. Medium says the technology can amplify human intelligence, while acknowledging the risk of low-quality content and confusion between human-written and AI-generated text. To improve transparency, TK lets authors publish notes about how the tool was used, including prompts and source material. The app also has a social layer for sharing drafts and receiving feedback; a desktop version is planned.

Sources: TechCrunch

Read the original story ↗

Disney and Adobe announced a multiyear partnership to develop proprietary, artist-led generative AI models tailored to Disney properties. According to Adobe, the models will be trained on specific knowledge about the company’s characters, environments and visual styles to generate commercially safe, production-ready outputs.

Disney will receive early access to Adobe’s latest technologies to support workflows for audiovisual productions, theme parks, its cruise line and ESPN graphics and content. The partnership will also explore emerging technologies, marketing activations and event collaborations, with a stated focus on responsible AI development and protecting Disney’s creators. The Verge reports that Disney’s Imagineering R&D division already uses the Firefly platform for theme park development.

Sources: The Verge, Tecnoblog

Read the original story ↗

When announcing the second release candidate for Linux 7.4, Linus Torvalds said that, at least in the eBPF area, fixes may be arriving faster than new problem reports, citing Alexei Starovoitov, one of the technology’s maintainers. Torvalds added that this would be positive “if true.”

eBPF allows software to run in a controlled way inside the kernel. Torvalds considers using AI to find bugs the “new normal” in Linux, but the excess reports—including low-priority issues or problems that have already been fixed—could still overwhelm the team. Linux 7.4 is expected in December 2026.

Sources: Tecnoblog

Read the original story ↗

Employment in programming and IT reached 563,400 workers in the third quarter, the highest level in the INE series. Data cited by Xataka and attributed to Business Insider put technology employment at 682,400 workers, up 2.3% year over year; programming and consulting added 31,200 workers in the period after losing 23,400 in the first quarter. Telecommunications moved in the opposite direction, losing 6,100 jobs and falling to 105,800 workers, its lowest level since 2008.

Xataka says the figures are not conclusive on AI’s impact: the 2026 annual average of employed people was 668,600, 2.8% below 2025. The first digital talent report from Mobile World Capital, meanwhile, found that demand for junior professionals fell 25.2% between 2024 and 2025, while demand for senior profiles rose 15.4%. The figure measures job postings, not employment, and Rafael Doménech of BBVA Research describes the situation as a reassignment of tasks, wages and entry opportunities.

Sources: Xataka

Read the original story ↗

According to Wired, Nvidia CEO Jensen Huang began working with crisis communications firm Heller at least as early as October 2025, months before admitting to an extramarital relationship with Alice Huang. The firm was reportedly instructed to handle any press inquiries about Huang, whether personal or professional. Founded by Risa Heller, the firm is known for working on personal controversies and also handles corporate clients.

The New York Times had linked Alice to Nvidia’s dealings with Megaspeed and reported that US intelligence officials identified her as a spy working for the Chinese government. Huang apologized to his family and employees, and Nvidia’s board announced an independent investigation. The board said it had no reason to believe that confidential company information or national security information had been compromised.

Sources: Wired

Read the original story ↗

Jonathan Kanter, the former head of the US Justice Department’s Antitrust Division, and Doha Mekki, the division’s former acting chief, launched FluxCap. The startup sells a platform built with proprietary technology and large language models for legal research and analysis, initially focused on antitrust cases and planned to expand into other practice areas. Kanter says the tool can identify weaknesses in legal arguments, sift through large volumes of documents and find signals of potentially anticompetitive conduct.

The team includes former Justice Department staffers and former Palantir engineers. Antitrust boutique Kressin Powers is among FluxCap’s first customers, and Kanter says the company is in talks with public and private enforcers. He says the platform can give a small team the capacity of “20-plus people” working around the clock, acting as a force multiplier for investigators and lawyers.

Sources: The Verge

Read the original story ↗

About 600,000 Nvidia H200 chips were reportedly being stored in Hong Kong awaiting shipment to mainland China, Reuters reported, as cited by Xataka. Two people familiar with the matter told Reuters that large companies may have obtained about 10,000 GPUs each, but the exact volume is uncertain because of the gray market, intermediaries and indirect routes used to bypass trade restrictions.

The U.S. authorized H200 sales to selected Chinese customers, and Reuters reported that about 10 companies gained access to the chips in May, including Alibaba, Tencent and ByteDance. However, The Information reported that China limited approvals to fewer than 200,000 units and directed companies to prioritize domestic processors for inference, widening the gap between the estimated Hong Kong stockpile and officially authorized purchases.

Sources: Xataka

Read the original story ↗

At the Conference on Robot Learning (CoRL) in Austin, Texas, NVIDIA announced GR00T 2, an open world action model built on NVIDIA Cosmos, along with simulation tools, benchmarks and workflows for agentic robotics. The company says the model completed at least once 40 of 50 evaluated tasks without task-specific demonstrations and achieved a 71% average success rate with one demonstration per task, rising to 95% with five. The system can also transfer skills between different robots and help AI agents execute sequences and recover from errors.

Two NVIDIA papers are among CoRL’s six Best Paper Award finalists: GRAIL, about generating 3D assets and simulation-ready scenes, and RoboTTT, which expands the context of visuomotor policies by three orders of magnitude without increasing inference latency, according to the company. NVIDIA says organizations including Nexuni, Foxconn and Siemens are already testing GR00T 2 for robotic laundry handling and industrial automation.

Sources: NVIDIA

Read the original story ↗

Automattic, the owner of WordPress.com, launched Spacefast, a platform for deploying websites, apps and services built with vibe coding. The service provides infrastructure including databases, authentication, file uploads, email and analytics, and can connect to agents and chatbots such as ChatGPT, Claude, Grok Bot and Cursor, or to a Git repository. The platform competes with services such as Lovable and Emergent, but lets users choose which AI tools they use. The product is led by Batuhan İçöz, who initially developed it during an internal company project; Automattic soft-launched it in September to gather feedback.

According to İçöz, people and organizations have already used Spacefast for internal mockups, presentations and offer-related websites. Automattic also plans to draw on existing products such as WordPress, web analytics and WooCommerce to serve enterprises seeking to reduce tool sprawl. The executive acknowledged that some users may leave after trying the service, and said the priority is to make building simple and the results better than they would be without the platform.

Sources: TechCrunch

Read the original story ↗

NVIDIA announced an investment in Point2, a South Korean AI interconnect company, and said it also plans to invest in Motif Technologies, RLWRLD and Upstage. The announcements were made at the MSIT-NVIDIA AI Startup Partnership Forum, held at COEX with South Korea’s Ministry of Science and ICT and attended by Vice Minister Ryu Je-myung.

According to NVIDIA, more than 20 startups took part, spanning foundation models, physical AI, AI infrastructure and AI semiconductors. Motif develops models for agents, RLWRLD builds robotics models for real-world tasks, and Upstage works on language models and document AI tools. NVIDIA also said its NVIDIA Inception program has more than 1,000 Korea-based members.

Sources: NVIDIA

Read the original story ↗

Platforms including KGen, Hub Data, Mimix AI, Flambra, Crowtado and Claru offer tasks in which users record videos of activities such as washing dishes, sweeping, cooking and shopping. Some participants interviewed by g1 say they received up to R$5,500 in a month; Yasmin Prado, 33, says she made about R$2,000 in September and that the activity has become her main source of income. Part of the earnings may come from commissions for referrals.

As the work has grown more popular, users report that payments have fallen and that they were banned for allegedly violating platform rules. The platforms do not disclose which companies buy the videos or explain in detail how they are used for training. Requirements include being at least 18, recording through one’s own account and avoiding faces, documents, screens, photos, vehicle license plates and outdoor areas. Participants must also follow rules on framing, lighting, duration and natural movements.

Sources: g1

Read the original story ↗

An unusual coalition is trying to stop Google from building a data center campus in Thurrock, east of London. The group includes conservation organization Buglife, local wildlife groups, nonprofit Foxglove, and the Thurrock Hammers motorcycle club. They cite the possible loss of habitat for rare plants and animals, pollution, and the loss of a racing track on the site; the club also wants to race there again. A petition opposing the project has collected almost 55,000 signatures, according to Wired.

Google acquired the site through a subsidiary in 2023 and applied for permission to build the campus. The opposition group says the development could produce up to 570,000 tonnes of emissions a year. Google says its plans prioritize biodiversity and emissions reduction, and that the objections are based on misconceptions about the project. The dispute comes amid protests against data centers elsewhere in Europe, according to Wired.

Sources: Wired

Read the original story ↗

Stephen Zhu, DiDi’s senior vice president, told Tecnoblog that the company recorded an “insignificant” number of rides requested by external artificial intelligence agents, such as those in ChatGPT or Claude. He questioned what additional value those agents provide, since customers are accustomed to using DiDi’s app — 99 in Brazil. The previous week, Uber had said that only four rides had been requested in this way.

DiDi has its own travel assistant, XiaoDi, which accepts voice or text commands and selects from 90 different service options. According to Zhu, it already handles requests such as a car that can accommodate a dog, but the company does not classify those rides as requests made by external AI agents. DiDi operates in 15 countries and says it has more than 749 million annual active users worldwide.

Sources: Tecnoblog

Read the original story ↗

Nvidia opened an independent investigation into Jensen Huang after reports linked his relationship with Alice Huang, a former executive at data center company Megaspeed, to a possible Chinese intelligence operation. According to reports cited by The Verge, the BBC, and Xataka, US officials suspect Alice Huang worked for Chinese intelligence and passed information to Beijing; the allegations have not been proven, and China rejected them. Huang admitted being unfaithful to his wife but did not name Alice Huang in his message to employees. Nvidia says it has no reason to believe company secrets or US national security were compromised. The BBC reported that the company hired Wachtell, Lipton, Rosen & Katz, led by attorney Joon Kim, to conduct the investigation; the FBI is also investigating. Donald Trump said Huang told him he did not know Alice Huang was a spy and that he believed him.

Sources: The Verge, TechCrunch, g1, Xataka, BBC, Ars Technica, Wired

Read the original story ↗

According to Bloomberg’s Mark Gurman, as reported by The Verge, CEO John Ternus is pressuring Craig Federighi to accelerate Siri’s AI roadmap and build an agent that can compete with OpenAI’s Dots and Meta’s Muse. Apple is preparing an update that would give the assistant more detailed control over its own apps, such as cropping and rotating images in Photos and organizing Safari tabs; those features are expected in iOS 27.3.

Integration with third-party apps is expected later, in the spring, enabling actions such as placing orders through Grubhub or booking restaurants through OpenTable. Gurman also says Apple wants Siri to track goals and carry out multistep tasks without the user initiating every step—an evolution that could arrive as early as next year.

Sources: The Verge

Read the original story ↗

Rights holders for Elvis Presley’s music and the publisher of some of Eminem’s biggest records have sued Suno, claiming that the company generates music and lyrics that imitate the artists. In the Presley-related suit, lawyers argue that Suno violated Tennessee’s Ensuring Likeness Voice and Image Security (ELVIS) Act, which bars AI companies from cloning a singer’s voice. The lawsuits add to cases brought by Sony, UMG, Jason Isbell and Reservoir Media, among others.

Suno spokesperson Cat McCormack told The Verge that the company exists to help users create new, original music, not to trade on artists’ names or likenesses. She said Suno does not allow prompts containing specific artists’ names and songs and uses third-party providers to screen uploaded audio files and lyrics, while continuing to strengthen its safeguards.

Sources: The Verge

Read the original story ↗

Wired describes world models as systems that receive images, recreate environments in 3D and simulate physics, cause and effect. According to Oliver Cameron, CEO of the lab Odyssey, they predict the world’s next visual state; Brian McClendon, chief technology officer at Niantic Spatial, says millions of simulations could train robots in specific spaces without physical movement. Wayve, meanwhile, uses car-camera footage to recreate hazardous roads and conditions that are difficult to test in real life.

The technology is also being explored in video games, with investment from Tencent, and may gain ground in home robots and autonomous cars—Wired identifies 2027 as a possible important year for that expansion. Development requires real-world data: the report says workers are already recording household tasks on video to sell through apps, generally for low payments.

Sources: Wired

Read the original story ↗

AI safety researchers are studying reverse game theory, also known as mechanism design, as an alternative to investigating models’ internal workings or directly training them to follow human values. The approach seeks to create rules and incentives that lead agents to behave as desired, even when their intentions are not transparent.

According to MIT Technology Review Brasil, a preprint by MIT economist Stephen Morris, with researchers from Yale and Columbia, proposes rewarding models for their ability to predict another model’s responses rather than evaluating performance alone. This could make it harder for an agent to pretend to be less capable in order to avoid safety mechanisms. The paper also suggests competitions between agents and pairs with opposing biases, using agreement between their answers as an incentive.

Sources: MIT Technology Review Brasil

Read the original story ↗

NYT Cooking has added an iOS-only beta AI feature for subscribers that lets users request ingredient substitutions, quantity adjustments, measurement conversions, and changes for different equipment or dietary needs. The tool creates a new version without altering the original recipe and visibly highlights its changes; requests considered unsuitable may receive a refusal and a more appropriate suggestion. The Verge found that the assistant was also trained on user comments and offered practical substitutions, although it nearly suggested using half the required butter in a cake.

WIRED likewise found the feature surprisingly effective in testing, but reported a negative user reaction: after four days and 172 almost exclusively critical comments on a post about the feature, commenting was turned off. The magazine said the system adapted recipes to available ingredients and equipment and redirected highly unsuitable requests to more appropriate recipes.

Sources: The Verge, Wired

Read the original story ↗

In an interview with The Verge, Alex Stamos, Cognition’s chief security officer and a cybersecurity expert, said private and state actors, including Russia and North Korea, are using AI to find and exploit vulnerabilities at the same time. He said the multibillion-dollar cybersecurity industry is facing a succession of breaches and may enter a “dark age” while it adapts to the new reality.

Stamos cited an attack chain involving Fireworks and Cursor: as described in the interview, API keys stolen from Fireworks enabled an intrusion into Cursor, which then distributed a malicious plugin to other companies. The plugin reportedly reached 5,000 companies, creating 5,000 potential breaches. He described the situation as a continuous sequence of breaches across the technology sector, with attackers chaining compromises to expand their reach.

Sources: The Verge

Read the original story ↗