Tech·6,172 stories
OpenAI has introduced a new memory architecture for ChatGPT that is more capable and compute‑efficient. The update also adds a summary page that lets users review and control what the model remembers,
TechCrunch is bringing Startup Battlefield back to Sydney on August 19, 2026, in partnership with Stripe, with eight Australian startups set to pitch live for a guaranteed entry into TechCrunch Disrup
Poke, a startup that lets users access AI agents via text message, has been approved by Apple as the first AI agent on its Messages for Business platform. The approval opens a new channel for AI‑drive
Anthropic has placed about six engineers inside the NSA to assist the agency in deploying its Mythos AI for offensive cyber operations. The move coincides with a legal dispute between Anthropic and th
Anthropic warns that AI systems may soon be capable of recursively designing and training more advanced successors without human input, accelerating progress in the near future and prompting calls for
Canada announced a new artificial intelligence strategy on Thursday, promising to generate 250,000 jobs by 2031. The plan includes a CA$500 million technology fund to back domestic AI startups.
An unauthorized group claimed to have accessed Anthropic’s Claude Mythos model within hours of its technical preview, highlighting the security challenges of deploying agentic AI in U.S. defense netwo
The Verge is running a sweepstakes that offers a Nomatic Travel Pack packed with over $800 of tech. The contest runs from 10 a.m. ET on June 4 to 10 a.m. ET on June 30, with the winner selected and no
Google has released Gemma 4 12B, an 11.95B-parameter open multimodal model that processes text, images, and audio in a unified encoder-free architecture and runs locally on laptops with 16GB of RAM un
A surge of commentary across The Atlantic, The Verge, and The Overspill converges on the view that large language models lack true consciousness, while pointing out that companies like Anthropic lean
Google released Gemma 4 12B, an open-source 11.95B-parameter multimodal AI model that processes text, images, and audio and runs locally on laptops with 16GB of RAM.
Leading AI and biotech executives have urged Congress to make mandatory screening of synthetic DNA and RNA orders, warning that AI‑driven design could lower barriers to creating harmful pathogens and
The Download, a weekday tech newsletter, reports that courts are overwhelmed by AI-generated lawsuits. Colorado federal magistrate judge Maritza Braswell spends most of her days sifting through these
AccessOwl, a profitable Y Combinator-backed startup founded in Berlin and San Francisco, is hiring a Senior Software Engineer to add 300 SaaS integrations this year using TypeScript, AI, and Playwrigh
Nvidia is expanding its AI ambitions beyond data centres by teaming with Dell, Lenovo and Asus to ship Windows laptops with integrated AI capabilities in the third quarter of 2026.
Waymo is teaming with B2U Storage Solutions to give its retired robotaxi lithium‑ion packs a second life as grid‑connected energy storage in California and Texas. The repurposed batteries will store e
Acclaimed sci-fi author Ted Chiang published a new Atlantic essay arguing that LLM conversations are sophisticated sentence-completion, not evidence of machine consciousness — and the internet has opi
Amazon unveiled an upgraded Proteus warehouse robot that responds to spoken commands instead of code, letting workers direct it conversationally. The autonomous system, now operable across entire fulf
Google unveiled the AI Edge Gallery for macOS, letting users run Gemini models directly on their Macs. The rollout also highlights the new Gemma 4 12B multimodal model and open‑source tools for local,
Microsoft announced at its Build conference that it has effectively ended its exclusive partnership with OpenAI, aiming to build AI capabilities in‑house. The company unveiled a new reasoning model, M
Curated tech stories, every morning. Free.
No spam. Unsubscribe anytime.