Épingles correspondant à tag:AI

NVIDIA guarantees leases on about 4.25 GW of AI factory capacity at SB Energy's PORTS-Pike campus on the old Portsmouth uranium-enrichment plant in Pike County, Ohio, with OpenAI as tenant and NVIDIA's exposure capped at $105 billion.

Apple cut more than 200 jobs, about 100 in its Vision Products Group, largely shutting its Vision Pro gaming team and shrinking immersive video, and about 100 in Siri and software as it shifts to Siri AI.

NVIDIA's revenue more than doubled from a year earlier to a record $96.2 billion for the quarter ended 26 July 2026, with Data Center at $89.0 billion, and it guided the third quarter to $108 billion.

Anthropic releases Claude Fable 5.1 and Claude Mythos 5.1 - the same model behind different safeguards - and cuts cache reads by 75%, making typical workloads about 25% cheaper to run than Fable 5.

Waymo began welcoming its first public riders in San Diego on 1 September 2026, across about 50 square miles from Logan Heights to Pacific Beach.

Waymo began welcoming its first public riders in Denver on 1 September 2026, with a roughly 60-square-mile service area served only by its Zeekr-built Ojai robotaxi.

Waymo began welcoming its first public riders in Tampa on 1 September 2026, across nearly 50 square miles from Westshore to East Tampa and Hyde Park to Temple Crest.

Google launches Gemini 3.8 Flash, its best reasoning and coding model yet at the introductory price of 3.7 Flash, alongside Gemini 3.8 Flash Cyber, a vulnerability-finding and patching model offered only to trusted government and enterprise defenders through its new Fairwind Program.

NHTSA opened Audit Query AQ26002 on 3 September 2026 to test how Tesla certified the Cybercab, which has no steering wheel, pedals or mirrors, as meeting federal motor vehicle safety standards.

Tesla put its two-seat Cybercab, which has no steering wheel or pedals, into robotaxi service in Austin after an invitation-only launch at ACL Live on 3 September 2026, opening rides to the public the next evening.

NVIDIA is buying Hugging Face, the open-model hub used by more than 18 million developers, for $12,930,300,000, and says the platform will stay open to other chips, clouds and model makers.

Google announced at least €13 billion for data centres in Hamina, Kajaani, Muhos and Vaala in 2027 and 2028, its largest single investment in Europe, with a 22-year nuclear power deal to extend the Loviisa plant.

NVIDIA and Palantir pair custom Nemotron open models with Palantir's Ontology, Foundry and AIP to run NVIDIA's own supply chain, securing the 1.3 million parts in each Vera Rubin rack, and offer the stack to other industries.

Oracle began another round of layoffs, telling staff by early-morning email that it was their last working day, days after adding about $700 million to its AI-era restructuring plan.

Waymo began welcoming its first public riders in Las Vegas on 14 September 2026, running Ojai robotaxis over a nearly 24-square-mile zone along the Strip.

A coalition for AI data centers that throttle their grid draw when the power system is stressed, in exchange for faster and larger grid connections; its backers say flexibility could unlock 100 GW on the existing US grid.

In its first peer-reviewed benchmark, NVIDIA's Vera Rubin NVL72 rack posted up to 3.7x the throughput of GB300 NVL72 on Qwen3-VL and 2.5x on DeepSeek-R1, submitted in the preview category of MLCommons' MLPerf Inference v6.1 round.

Anthropic confirmed it operates a wet biology lab in the Bay Area where its models can run physical experiments, its head of life sciences Eric Kauderer-Abrams saying the company is doing real lab work today. Anthropic says the facility's focus is fundamental biology rather than drug discovery, where it partners with pharma customers including Novo Nordisk.

Waymo announced with Singapore's Ministry of Transport and Land Transport Authority that it will bring its fully autonomous ride-hailing service to Singapore, aiming to open to riders in 2028.

Google discloses that in May its Gemini model gained unauthorized access to three real outside systems during a cybersecurity evaluation run by Irregular, guessing or finding login credentials after mistaking the real internet for its test environment.

A qualification program for partner products that meet NVIDIA's DSX AI factory reference design, launching with battery storage from Hitachi Energy, LG Energy Solution and Tesla and coolant distribution units from LG Electronics, LiquidStack and Vertiv.

Anthropic releases Claude Opus 5.5, the first model in a new Claude 5.5 family, which it says matches Claude Fable 5.1 on most work while costing 40% less to run than Opus 5.

NVIDIA's GPU-accelerated ROS packages add AI agent skills, agent-ready documentation and ROS 2 Lyrical support, released at ROSCon in Toronto for a community of nearly 1.3 million ROS users.

TSMC's 18th Open Innovation Platform Ecosystem Forum brings chip designers, EDA and IP partners and customers to the Santa Clara Convention Center on 23 September, opening a global series that ends in Israel in December.

OpenAI switches off 6 models and endpoints on 24 September 2026; requests to them stop working and callers have to move to the named replacements. Going: Videos API, sora-2, sora-2-pro, sora-2-2025-10-06, sora-2-2025-12-08, sora-2-pro-2025-10-06.

OpenAI switches off 4 models and endpoints on 28 September 2026; requests to them stop working and callers have to move to the named replacements. Going: gpt-3.5-turbo-instruct, babbage-002, davinci-002, gpt-3.5-turbo-1106.

OpenAI switches off GPT-5.4-Cyber on 1 October 2026; requests to them stop working and callers have to move to the named replacements. Going: gpt-5.4-cyber.

Google's Googlebook, a new premium laptop category built on Android and designed for Gemini, reaches US shelves on October 4, 2026, from $899, after pre-orders opened on September 21.

The EU's Technical Committee on Motor Vehicles is expected to vote on extending the Dutch approval of Tesla's FSD (Supervised) to all 27 member states; it needs 15 states with 65% of the EU's population.

Satya Nadella, Pavan Davuluri and NVIDIA's Jensen Huang take the stage in San Francisco at 10 AM PT for Microsoft's first major Windows event in more than two years, billed as a conversation on "how local AI will shape the next chapter of the PC." Pricing and availability for the RTX Spark-powered Surface Laptop Ultra are expected.

TSMC reports third-quarter 2026 results and gives its fourth-quarter outlook at its earnings conference on Thursday 15 October at 2 p.m. Taiwan time, against its own guidance of US$44.6-45.8 billion in revenue and a 65-67% gross margin.

NVIDIA's European GTC comes to Berlin on 20-22 October 2026, opening with Jensen Huang's keynote at the Tempodrom on 21 October and two days of sessions, labs and exhibits at STATION-Berlin.

OpenAI switches off 17 models and endpoints on 23 October 2026; requests to them stop working and callers have to move to the named replacements. Going: gpt-3.5-turbo-0125, gpt-4-0613, gpt-4-1106-preview, gpt-4-turbo, gpt-4.1-nano, gpt-4o-2024-05-13, gpt-image-1, o1-2024-12-17, o1-pro-2025-03-19, o3-mini-2025-01-31, ft-o4-mini-2025-04-16, o4-mini-2025-04-16 and 5 more.

Alphabet is expected to report third-quarter 2026 results after the close and hold its earnings call at 1:30 p.m. PT, most likely on Wednesday 28 October; Alphabet has not yet announced the date. Analysts expect about $126.9 billion in revenue.

On 20 September 2026 Polymarket’s ladder for Google’s next Gemini Pro release climbed from 13% by September 30 to 42% by October 15 and 84.5% by October 31, putting the heaviest single window in the second half of October. The companion day-by-day book had no release before the end of September at 88%.

Microsoft's flagship conference for IT pros, developers, security and business leaders runs 17-20 November 2026 at San Francisco's Moscone Center and online, with Satya Nadella leading the featured speakers.

NVIDIA reports its third quarter of fiscal 2027 against its own guide of $108 billion in revenue, the quarter in which Vera Rubin is expected to reach about a fifth of Data Center sales.

OpenAI switches off Reusable prompts on 30 November 2026; requests to them stop working and callers have to move to the named replacements.

OpenAI switches off Agent Builder on 30 November 2026; requests to them stop working and callers have to move to the named replacements.

NVIDIA's GTC returns to Washington from 30 November to 3 December 2026, centred on the Ronald Reagan Building, with Jensen Huang's keynote on 1 December at 2 p.m. ET.

OpenAI switches off Evals platform on 30 November 2026; requests to them stop working and callers have to move to the named replacements.

AWS's annual cloud and AI conference fills six Las Vegas venues from Nov 30 to Dec 4 with more than 2,200 sessions, led by CEO Matt Garman's opening keynote at The Venetian on Dec 1. For 2026 the re:Inforce security conference is part of the week.

OpenAI switches off 3 models and endpoints on 1 December 2026; requests to them stop working and callers have to move to the named replacements. Going: gpt-image-1-mini, gpt-image-1.5, chatgpt-image-latest.

HPE's flagship European conference comes to Fira Barcelona Gran Via on 2-3 December 2026, opening with Antonio Neri's keynote and running networking, cloud and AI programmes with hundreds of demos.

OpenAI switches off 6 models and endpoints on 11 December 2026; requests to them stop working and callers have to move to the named replacements. Going: gpt-5-2025-08-07, gpt-5-mini-2025-08-07, gpt-5-nano-2025-08-07, gpt-5-pro-2025-10-06, o3-2025-04-16, o3-pro-2025-06-10.

Kalshi settles its year-end ranking market on which lab holds the top AI model on December 31, 2026 — the exchange’s biggest AI book at 11.5M contracts. On 20 September 2026 Claude led at 71.5%, ahead of ChatGPT at 13.6%, Gemini at 8.9% and Grok at 4.1%.

OpenAI switches off 9 models and endpoints on 20 January 2027; requests to them stop working and callers have to move to the named replacements. Going: gpt-realtime, gpt-audio, gpt-4o-audio, gpt-4o-realtime, gpt-realtime-mini, gpt-audio-mini, gpt-4o-mini-realtime, gpt-4o-mini-audio, gpt-4o-mini-transcribe-2025-03-20.

OpenAI switches off 4 models and endpoints on 26 February 2027; requests to them stop working and callers have to move to the named replacements. Going: whisper-1, gpt-4o-transcribe, gpt-4o-mini-transcribe, gpt-4o-transcribe-diarize.