AI news for builders and product teamsUpdated Nov 17, 2026, 21:01 UTC
AI news, straight from the source.
Releases, research, and ideas for developers and product teams. Updated every hour.
Across the feeds
Newest first
Adobe's Lightroom app for iOS, iPadOS and Android can now open closed eyes in photos with a single tap using generative AI, surfacing an Open Eyes Quick Action when it detects shut eyes. Adobe announced the feature yesterday.

A Ben's Bites contributor built a single dashboard bringing todos, email, CRM and domains into one visual workspace where a human and their agents can work side by side, and published a video plus a write-up with copy-paste agent instructions so others can replicate it.

Eka Robotics raised a $150 million Series A led by a16z and F-Prime at a $1.45 billion post-money valuation, on the strength of a manipulation system that it says picks up objects more than 95% of the time.

Meta has suspended Muse's conversation-sharing feature after Claroc researchers found 6,800 share links by enumerating a predictable subdomain pattern, then opened a few hundred and found confidential business and personal material.

Mistral AI launched Voxtral 3, a family of speech models plus a voice agent SDK that runs listening, reasoning, tool calls, and speech in one real-time pipeline across 50+ languages. It is available via Mistral's API or self-deployment, with transcription from $0.0035 per minute and TTS at $13 per million characters.

The US Department of Homeland Security has barred federal civilian agencies from using DeepSeek AI technology, requiring them to remove existing products and blocking new purchases, the Wall Street Journal reports.

Yelp is adding a "Check for Me" feature that lets its app place an AI voice call to a restaurant on a user's behalf to ask about wait times, then texts the answer back. The update also adds OpenTable booking, persistent chat preferences, custom lists, and a $30-per-month AI lead-reply product for businesses.

Otter.ai has raised $320 million from General Catalyst's Customer Value Fund, a vehicle that puts money into startups in exchange for a share of marketing-spend results rather than equity. The company says the capital will go toward enterprise growth, go-to-market, international expansion and product work.

NVIDIA says AI cloud providers Nebius, Nscale and Verda are building infrastructure on its Vera CPU, with server makers Dell, HPE, Lenovo, Supermicro and Bull also building Vera-based systems. Vera is positioned for agentic workloads, especially the fast sandbox startups and data processing that agents rely on.

An independent Schlotzsky's and Cinnabon franchisee says it reconciles cash in 94% less time using ChatGPT Work, and cut menu repricing time by 95% while building a workflow the release says could free up $100K in cash flow.

Samsung's Pyeongtaek semiconductor complex in South Korea has three fabs running around the clock, a fourth coming online, and a fifth under construction after work paused in 2024 during a memory demand slump and restarted as AI demand surged.

DeepSeek's V4.1 Flash 1017 is a 552B-parameter, 16B-active open-weight MoE model under an MIT license with a 1M-token context window, priced unchanged at $0.30/$1.20 per million tokens; Artificial Analysis scores it 45 on its Intelligence Index, tied for second among open-weight models.

DeepSeek's V4.1 Flash 1017, a 552B-parameter open-weight MoE model with a 1M-token context and MIT license, lands near the top of open-weight rankings at $0.30/$1.20 per million tokens, and MiniMax's M3.1 Flash and Inception's diffusion model Mercury 3 also shipped.

NVIDIA says some Barcelona Supercomputing Center AI Factory services are now reachable through the NVIDIA Inception portal, and three Inception members were named AI Champions at the programme's pilot demo day.

New York Mayor Zohran Mamdani told an AI Now Institute event at Webster Hall that New Yorkers should reject the idea of an inevitable AI takeover and demand accountability, pointing to what he called a federal abdication on AI regulation.

AnyLanguageModel 1.0 is out, a Swift package that keeps Apple's Foundation Models API but lets apps swap in local llama.cpp, MLX or Core ML models, or hosted providers like OpenAI, Anthropic, Gemini and Ollama, via a single import change. swift-huggingface 1.0 and swift-xet 1.0 ship alongside it.

Hugging Face announced TorchTPU, a native PyTorch backend that makes Google TPUs a regular device, so Transformers, Diffusers and Kernels now run on TPU by passing device_map="tpu" without torch_xla or Optimum TPU.

OpenAI has agreed to acquire Orchestra Bio, aiming to apply AI-powered operations to life sciences R&D and help biopharma teams speed up drug development.

Simon Willison highlights Jeremy Scheff's essay on how coding agents erode a core satisfaction of programming careers: getting paid to solve puzzles that ordinary people could not. He argues the loss is real for those who enjoyed direct problem-solving, while those focused on building systems benefit from LLMs.

Anthropic and AI infrastructure provider Nscale are reportedly close to announcing a second $45 billion compute deal, about a month after announcing a first one of the same size, according to Bloomberg. Talks are still ongoing and terms could change.