Tech·4,615 stories
Alibaba announced that its Qwen AI large‑language model will be built into the infotainment systems of selected BYD and Volkswagen vehicles. The integration lets drivers place voice‑activated orders f
Chinese AI startup DeepSeek unveiled a preview of its V4 model, asserting it can match US giants like Anthropic, Google, and OpenAI in coding and general tasks. The launch, built for Huawei chips, arr
OpenAI touts ChatGPT as a helpful budgeting aid, yet experts warn its confident tone masks lingering hallucinations and sycophancy that can mislead users. As Anthropic, Google, and OpenAI race to embe
An interactive HTML guide translates Andrej Karpathy's technical lecture into a walkthrough of how ChatGPT-style models are built — from filtering 15 trillion tokens of internet text to post-training
Tracxn’s data shows a dramatic collapse in global edtech, with funding falling from $16.7 B in 2021 to $2.6 B in 2025 and startup launches plunging from 7,178 to just 645. Venture capital is abandonin
DeepSeek released preview versions of its V4 Pro (1.6T parameters, 1M token context) and V4 Flash models, posting both to Hugging Face roughly a year after its earlier breakthrough disrupted Silicon V
DeepSeek launched preview versions of its V4 Pro and V4 Flash AI models, positioning them as high‑performing, cost‑effective alternatives that challenge US rivals while still lagging top‑tier performa
OpenAI launched GPT-5.5, a new AI model designed to autonomously handle complex tasks with minimal guidance, and said it will power the company's upcoming "super app."
DeepSeek has released two new AI models, V4 Pro and V4 Flash, positioning itself as a competitive force in the global AI race with efficient, high-performance open-source models.
Anthropic expanded Claude's app connectors beyond workplace tools into personal services like Spotify, Uber, TurboTax, and Instacart, making the new integrations available to all users with mobile in
Sierra, the Bret Taylor‑founded AI customer‑service platform, has bought French workflow‑AI startup Fragment, adding European talent to its expanding portfolio after recent deals in Japan and voice AI
Texas Instruments’ shares surged 19% after the chipmaker issued upbeat Q2 guidance, driven by soaring demand for its analog chips in AI data centers. The rally mirrors a broader chip boom, as Intel re
OpenAI launched GPT-5.5, an AI model designed to complete complex tasks with minimal guidance, and confirmed the model will power its upcoming agentic 'super app.'
Meta announced it will lay off about 8,000 employees, roughly 10% of its workforce, as part of an AI‑efficiency drive amid soaring AI costs and a projected 83% plunge in free cash flow.
Anthropic’s Claude Desktop for macOS silently creates a pre‑authorized Native Messaging manifest that grants its browser extensions privileged access to a local binary without user consent, raising pr
OpenAI launched GPT-5.5, pricing the new model at double GPT-5.4's rates — $5 per million input tokens and $30 per million output tokens — while positioning it as the engine behind an upcoming 'super
Anthropic’s flagship Claude Mythos, touted as a cyber‑secure AI too dangerous for public release, was quietly accessed by a small group of unauthorized users. The breach, traced to an educated guess u
DeepSeek released preview versions of its V4 Pro and V4 Flash AI models on April 24, 2026, roughly a year after its previous release shook Silicon Valley, with the company saying V4 Pro trails state-o
Microsoft is rolling out a new Agent Mode across Word, Excel, and PowerPoint, making its Copilot AI more interactive and default for Microsoft 365 Copilot, Premium, Personal, and Family users.
Apple has patched an iOS notification logging flaw that let the FBI forensically extract deleted Signal messages from a defendant's iPhone, closing a gap where encrypted communications persisted on-de
Curated tech stories, every morning. Free.
No spam. Unsubscribe anytime.