← Back to news

T

Techmeme RSS Feed

HSBC signs a Mistral deal to access Mistral's AI models and co-develop models for tasks like financial analysis, translation, onboarding clients, and AML checks (Ortenca Aliaj/Financial Times)

Omnicom's DM9 returned three Cannes awards after a North Carolina state senator and CNN Brazil found DM9 used AI to manipulate their content and used it for ads (Emmanuel Felton/Washington Post)

The AI boom has led to high demand and more pay for the construction workers who build data centers; a trade group estimates there's a shortage of ~439K workers (Te-Ping Chen/Wall Street Journal)

China's central bank reaffirmed its stance on crypto, calling virtual currency activity illegal and saying stablecoins fail KYC and anti-money-laundering rules (Reuters)

New York becomes the first US state to require that retailers disclose using algorithmic pricing tied to personal data; 10+ states are considering similar bills (Tim Balk/New York Times)

How Animaj, a startup in the Disney Accelerator Program, uses its generative AI tool to expedite the animation process, while claiming artists stay in control (Corinne Reichert/CNET)

US startups are increasingly adopting open-weight Chinese AI models, which are cheaper, more customizable, and sufficiently capable compared to frontier US ones (NBC News)

Q&A with ex-Intel CEO Pat Gelsinger on the "decay" at Intel, the Chip Act's terrible execution, the irony of Lip-Bu Tan broadly following his strategy, and more (Michael Acton/Financial Times)

How a Chinese entrepreneur amassed 10M+ IPv4 addresses, mostly from Africa, to lease them outside Africa, leaving African ISPs struggling to expand capacity (Alexandra Wexler/Wall Street Journal)

Anthropic says over 60% of its business customers use more than one Claude product, a trend it began noticing after Claude Code's rise in popularity (Richard Nieva/Forbes)

Sources: Databricks is raising $5B at a $134B valuation, expects sales to grow 55% in 2025 to $4.1B, and is operating at roughly breakeven (Cory Weinberg/The Information)

An in-depth look at TPUv7 Ironwood, the latest generation of Google's TPU, and how it positions Google as a serious challenger to Nvidia's AI chip dominance (Dylan Patel/SemiAnalysis)

South Korean authorities are investigating a data leak at e-commerce giant Coupang that exposed ~33.7M accounts; the country has a population of 51.7M (Sohee Kim/Bloomberg)

How David Sacks' work on AI and crypto in Trump's White House benefits his investments, those of his Silicon Valley friends, and the All-In podcast he co-hosts (New York Times)

Researchers unveil PropensityBench, a benchmark showing how stressors like shorter deadlines increase misbehavior in agentic AI models during task completion (Matthew Hutson/IEEE Spectrum)

OpenAI faces rising pressure from rivals three years after ChatGPT's debut; Similarweb says Gemini users chat longer per visit than ChatGPT and Claude users (Financial Times)

Alibaba Technical Report: Qwen3-VL beats GPT-5 and Gemini 2.5 Pro on visual tasks and has 100% accuracy on "needle-in-a-haystack" tests for 30-minute videos (Jonathan Kemper/The Decoder)

Taiwan raises its 2025 GDP growth forecast to 7.37%, up from 4.45% forecasted in August, the fastest in 15 years, as AI demand boosts its electronics exports (Kathrin Hille/Financial Times)

India's Department of Telecommunications directs WhatsApp and other messaging apps to implement SIM binding to ensure each account remains tied to an active SIM (Kamya Pandey/MediaNama)

An interview with Google DeepMind Nobel laureate John Jumper on the creative "off-label" uses for AlphaFold, combining AlphaFold's power with LLMs, and more (Will Douglas Heaven/MIT Technology Review)

A profile of Klay, whose app will let users remake songs using AI; Klay has raised ~$10M and is the first AI startup to close deals with the three major labels (Abram Brown/The Information)

Salesforce: online Black Friday sales grew 6% YoY to $79B globally, including $3B in US sales driven by AI agents; Adobe: US online sales rose 9.1% to $11.8B (Joan Verdon/Forbes)

Airbus issues an order to revert 6,000 A320-series planes to old software, after an incident revealed that intense solar radiation corrupted flight control data (Nadeem Badshah/The Guardian)

Sources: Micron plans to invest $9.6B in Japan to build a production facility for next-gen HBM memory chips beginning in 2026, with shipments expected in 2028 (Ryo Mukano/Nikkei Asia)

Pangram Labs: ~21% of the 75,800 peer reviews submitted for ICLR 2026, a major ML conference, were fully AI-generated, and 50%+ contained signs of AI use (Miryam Naddaf/Nature)

A look at Mexico City's Nuevo Polanco neighborhood, where Huawei, TikTok, and other Chinese firms have opened offices, fueling a growing Chinese tech community (Daniela Dib/Rest of World)

Unit 42 details how underground hacking forums advertise and sell custom, jailbroken, and open-source AI hacking tools such as WormGPT and KawaiiGPT (Derek B. Johnson/CyberScoop)

A look at ChatGPT's impact on US stocks three years after its release: AI mania revived Big Tech, making an already concentrated S&P 500 even more top heavy (Bloomberg)

Sunday Robotics, which debuted its Memo home robot on Nov. 19, hired at least 10 former Tesla employees, including engineers who worked on Autopilot and Optimus (Grace Kay/Business Insider)

Range, which uses AI to offer flat-fee financial planning services to mass affluent households, raised a $60M Series C led by Scale Venture Partners (Ryan Lawler/Axios)

An analysis of Google TPU v6e vs AMD MI300X vs Nvidia H100/B200: Nvidia achieves a ~5x tokens-per-dollar advantage over TPU v6e and 2x advantage over MI300X (@artificialanlys)

A profile of Paulina Borsook, as her 2000 book Cyberselfish, warning about Silicon Valley's love for "techno-libertarianism", finds a resurgence in interest (David Streitfeld/New York Times)

DeepSeek says its new DeepSeekMath-V2 model got gold-medal level status on the International Mathematical Olympiad 2025 and Chinese Mathematical Olympiad 2024 (Matthias Bastian/The Decoder)

SEC filing: CoinShares withdraws registrations for its XRP, Solana staking, and Litecoin ETFs to focus on higher-margin opportunities ahead of a US listing (Arasu Kannagi Basil/Reuters)

Prime Intellect, which builds infrastructure for decentralized AI training, debuts INTELLECT-3, a 106B open source MOE model it claims outperforms larger models (Prime Intellect)

Google withdraws its 2024 EU antitrust complaint against Microsoft over Azure licensing, after the EU launched a probe into Azure and AWS under the DMA (Edith Hancock/Wall Street Journal)

Indian e-commerce platform Meesho is set to launch an India IPO, seeking to raise $606M, targeting a post-issue valuation of ~$5.60B (Jagmeet Singh/TechCrunch)

Sources: Intel may start shipping a portion of Apple's lowest-end M chips as early as 2027; Apple is evaluating Intel's 18A process for the chip production (@mingchikuo)

French President Emmanuel Macron criticizes the EU for being too slow to act on probes into Big Tech companies, attributing the delay to US pressure against DSA (Politico)

Sources: Temasek is in talks to buy a ~5% stake in Quest Global, in a deal that values the engineering services company at $4.6B in the lead-up to its India IPO (Bloomberg)

MediaTek's stock posted its best week since 2002, up 22%, amid enthusiasm over Google's Gemini 3 and chip deals; MediaTek was reported as Google's TPU partner (Sangmi Cha/Bloomberg)

Meituan reports Q3 revenue up 2% to $13.4B and a $2.3B adjusted net loss, worse than $1.9B est., its first loss since 2022, amid a price war with JD and Alibaba (Luz Ding/Bloomberg)

Gravis Robotics, which adds cameras, sensors, and AI to heavyweight construction machines so they can operate autonomously, raised $23M led by IQ and Zacua (Beatrice Nolan/Fortune)

Global futures markets were disrupted for several hours on Friday after CME Group suffered an outage due to a cooling failure at a CyrusOne data center (Reuters)

The EC says it got notification from Apple that Apple Ads and Apple Maps likely meet DMA thresholds; Apple says the services should not be called "gatekeepers" (Reuters)

Stanford researchers develop a web-based tool that uses an LLM to downrank X posts with antagonistic language in a user's feed, to reduce "partisan animosity" (Stanford University)

Getty CEO Craig Peters says Getty may rethink its UK presence if the CMA blocks its Shutterstock deal, arguing the CMA is overlooking AI's impact on imaging (Daniel Thomas/Financial Times)

Nexperia warns customers face impending production halts and urges its Chinese unit to reengage, says it hadn't got "any meaningful response" despite outreach (Sarah Jacob/Bloomberg)

Australia passes a law requiring global streamers with 1M+ Australian subscribers to spend 10% of their local outlay or 7.5% of revenue on Australian originals (Jesse Whittock/Deadline)

OpenAI's data center partners are set to rack up nearly $100B in debt; sources say banks may lend another $38B to Oracle and Vantage to build more OpenAI sites (Financial Times)