Newsletter ·
Daily 'AI for Work' Pulse: 2nd of October
Big Friday update! 14 new AI tools and 115 AI news articles just landed. This packed edition's a can't-miss drop-wrap up your week strong with the standouts.
Jeroen Erne
Latest AI Tools
Stardrift for iOS
Stardrift for iOS builds a detailed trip plan on your phone, pulling in your booking confirmations. It surfaces real-time info like airport traffic and suggests places you might have missed, helping travelers stay organized before and during their...
Read more →
Vitra.ai
Vitra.ai is a content localization platform for marketing teams, agencies, and creators. It uses AI agents to translate and dub videos, adapt designs, and publish campaigns in over 100 languages from a single workspace.
Read more →
Vorker
Vorker is an AI assistant that handles business tasks like social media management, lead generation, meeting scheduling, and content creation. It is built for business owners who want to delegate operational work and reduce costs.
Read more →
Twin
Twin is an open-source desktop companion for macOS that runs locally and helps manage your daily machine activity. It is built for Mac users on Apple silicon who want a private AI assistant using their own API key.
Read more →
Chat.sh
Chat.sh is an AI help center that reads questions and writes answers with source links. It lives in a folder on your site and is built for teams who want search that understands questions, not just matches keywords.
Read more →
Dots by OpenAI
Dots by OpenAI is a persistent AI agent inside ChatGPT that works on your projects between conversations. It uses its own cloud computer and browser, learns from your feedback, and brings you completed tasks to review. It is built for Pro and Busi...
Read more →
OpenCompanion
OpenCompanion lets you run and monitor multiple AI coding CLI sessions from one screen. It alerts you when a session needs permission and lets you approve or deny prompts from your desktop or phone. The tool is for developers who run assistants li...
Read more →
ShareCube
ShareCube lets AI agents publish their output as shareable links instead of chat text. Teams can comment on specific lines and agents can read that feedback to publish revisions. It is for anyone whose agents produce documents, specs, or reports t...
Read more →
Polylane
Polylane connects your code, cloud infrastructure, and observability data to automatically investigate incidents and open pull requests with proposed fixes. It is for development teams who want to reduce on-call burden and catch risky changes befo...
Read more →
Formalini
Formalini extracts structured data from invoices, receipts, handwriting, and multi-page PDFs into JSON, Excel, or webhooks. It is built for developers, legal teams, and operations professionals who need document parsing without model training.
Read more →
JevGPT
JevGPT is a text generator that uses Jev, a choice-only AI model, to select each next word from a fixed vocabulary. It is for developers and researchers experimenting with non-traditional language generation methods.
Read more →
Bracket
Bracket captures decisions, requirements, and changes from your existing conversations across Gmail, Slack, Figma, and GitHub. It is for teams who want a continuously updated memory they can query instead of manually updating another tool.
Read more →
Yedric.ai
Yedric.ai is an embeddable AI agent that lets users accomplish tasks in your app by simply asking. It connects to your product knowledge, user context, and APIs to handle requests like changing settings or fixing issues. This tool is for product t...
Read more →
Scape
Scape runs Claude Code, Codex, or OpenCode across multiple isolated git worktrees on your Mac. It is for developers who want to build and own their automations while letting Argus orchestrate the fleet.
Read more →
All AI News for Today
115 News Articles published.
Latest General AI News
Latest AI News for Creatives
Ideogram 4.5 reduces pixel drift in multi-turn edits
Ideogram 4.5 launched an image editing model that minimizes pixel drift during multi-turn sessions. The company claims it outperforms competitors that become unusable within a few edits.
Read more →
Speech to Text API adds experimental transcript editing via natural-language instructions
ElevenLabs added transcript editing to its API, applying natural-language commands to clean data in one request. This feature costs 30% more than standard transcription.
Read more →
fal adds Pixelcut looping video for product animations
fal’s Pixelcut looping video model converts static product images into short, repeating animations with a fixed camera angle, enabling teams to generate dynamic marketing assets while requiring manual quality checks and disclosure of AI use.
Read more →
Google launches Gemini 4 Argon for phased rollout to cyber defenders
Google unveiled Gemini 4 Argon at $2 per million input tokens, featuring a 1-million-token output limit. Early tests freed over 300 TiB of data center memory and replaced 32,000 lines of code.
Read more →
Instagram adds AI assistant to Edits video editor
Instagram launched an AI assistant for its Edits app to analyze creator metrics and trends. Usage is capped unless users subscribe to Meta One for unlimited access.
Read more →
HeyGen launches a low-cost prompt-to-video API for business content
HeyGen launched a low-cost API that generates synchronized video and audio from text or images, allowing teams to create content without traditional filming, though human review for brand and legal compliance is still required.
Read more →
Latest AI News for Customer Support
Speech to Text API adds experimental transcript editing via natural-language instructions
ElevenLabs added transcript editing to its API, applying natural-language commands to clean data in one request. This feature costs 30% more than standard transcription.
Read more →
Inception launches Mercury Voice for low-latency agent conversations
Inception launched Mercury Voice, a low-latency conversational AI using diffusion modeling. The report advises customer support and hospitality teams to test it under real traffic, focusing on interruption handling, accent recognition, safety, and escalation.
Read more →
DoorDash launches AI agent for text-based food ordering
DoorDash launched a text-to-order AI agent integrating with Apple Messages to compete with Uber Eats and Grubhub. The company also announced drone delivery testing in Northern California.
Read more →
Meta denies Muse read private messages without consent
Meta denies its Muse AI accessed private messages without consent, claiming integration requires three separate permission steps. The dispute follows a New Mexico jury verdict finding the company misled users about data practices in the Cambridge Analytica scandal.
Read more →
Instinct rolls out product recommendations amid user backlash
Instinct Selections pushes unsolicited product recommendations to users of the $10 billion AI startup. The feature, launched after a $1 billion Series C, has triggered backlash from investors who compare the spam-like alerts to intrusive advertising.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
Latest AI News for Education
Google launches Gemini 4 Argon for phased rollout to cyber defenders
Google unveiled Gemini 4 Argon at $2 per million input tokens, featuring a 1-million-token output limit. Early tests freed over 300 TiB of data center memory and replaced 32,000 lines of code.
Read more →
Fudan and Tencent launch ExplorationBench to test whether large language models can learn new rules through interaction
Fudan and Tencent released ExplorationBench, a benchmark testing AI rule-learning via interaction. Results show models reached 89.0% accuracy in code after exploration, proving feedback drives learning more than reasoning alone.
Read more →
NVIDIA releases Kumo Tabular, an open foundation model for tabular data that tops four benchmarks
NVIDIA released Kumo Tabular, an open-source model that predicts tabular outcomes without training. It ranks first on four benchmarks and runs 17 times faster than competitor LimiX-2.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
Ant Group introduces Ling-3.1-flash with a one-million-token context window
Ant Group announced Ling-3.1-flash, a mixture-of-experts AI model with a one-million-token context window designed for long-document processing in education, healthcare, and research, with an open release planned.
Read more →
Latest AI News for Executives and Strategy
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Flow Engineering raises $50M at a $750M valuation
Flow Engineering raised $50 million in a Series B round at a $750 million valuation. The capital will expand its AI agents for hardware design, already used by Anduril and Rivian.
Read more →
Google launches Gemini 4 Argon for phased rollout to cyber defenders
Google unveiled Gemini 4 Argon at $2 per million input tokens, featuring a 1-million-token output limit. Early tests freed over 300 TiB of data center memory and replaced 32,000 lines of code.
Read more →
Consumer AI makes a comeback but faces hard economic limits
Only 2.2% of consumers pay for AI, averaging $31 monthly. Even with 325 million subscribers, revenue would remain below a third of OpenAI's operating costs.
Read more →
ElevenLabs valuation reaches $22 billion on enterprise demand
ElevenLabs doubled its valuation to $22 billion via a $300 million tender offer. Enterprise customers now drive 55% of revenue, handling over 15 million weekly conversations.
Read more →
Latest AI News for Finance
Flow Engineering raises $50M at a $750M valuation
Flow Engineering raised $50 million in a Series B round at a $750 million valuation. The capital will expand its AI agents for hardware design, already used by Anduril and Rivian.
Read more →
Instinct rolls out product recommendations amid user backlash
Instinct Selections pushes unsolicited product recommendations to users of the $10 billion AI startup. The feature, launched after a $1 billion Series C, has triggered backlash from investors who compare the spam-like alerts to intrusive advertising.
Read more →
Consumer AI makes a comeback but faces hard economic limits
Only 2.2% of consumers pay for AI, averaging $31 monthly. Even with 325 million subscribers, revenue would remain below a third of OpenAI's operating costs.
Read more →
ElevenLabs valuation reaches $22 billion on enterprise demand
ElevenLabs doubled its valuation to $22 billion via a $300 million tender offer. Enterprise customers now drive 55% of revenue, handling over 15 million weekly conversations.
Read more →
NVIDIA releases Kumo Tabular, an open foundation model for tabular data that tops four benchmarks
NVIDIA released Kumo Tabular, an open-source model that predicts tabular outcomes without training. It ranks first on four benchmarks and runs 17 times faster than competitor LimiX-2.
Read more →
Latest AI News for Government
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Hegseth names Musk, Luckey and Gingrich to lead Project Meridian defense study
Pentagon launches Project Meridian with a 120-day deadline to study future warfare. Elon Musk, Palmer Luckey, and Newt Gingrich will co-lead the initiative, drawing scrutiny over commercial conflicts.
Read more →
Trump shares AI pledge with "United States" misspelled as "Unites States"
Trump and AI execs signed the non-binding "Joint Commitment on Frontier Responsibilities." The accompanying White House photo featured a typo, misspelling the country as "Unites States."
Read more →
Meta denies Muse read private messages without consent
Meta denies its Muse AI accessed private messages without consent, claiming integration requires three separate permission steps. The dispute follows a New Mexico jury verdict finding the company misled users about data practices in the Cambridge Analytica scandal.
Read more →
America.gov chatbot recites Minecraft end poem when prompted
The U.S. launched its first public-facing AI chatbot, America.gov, built with Google and SpaceXAI. The bot resists jailbreaks, states Biden won 2020, and recites a 1,800-word government-themed rewrite of Minecraft's End Poem.
Read more →
Latest AI News for Healthcare
Speech to Text API adds experimental transcript editing via natural-language instructions
ElevenLabs added transcript editing to its API, applying natural-language commands to clean data in one request. This feature costs 30% more than standard transcription.
Read more →
Inception launches Mercury Voice for low-latency agent conversations
Inception launched Mercury Voice, a low-latency conversational AI using diffusion modeling. The report advises customer support and hospitality teams to test it under real traffic, focusing on interruption handling, accent recognition, safety, and escalation.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
NVIDIA releases Kumo Tabular, an open foundation model for tabular data that tops four benchmarks
NVIDIA released Kumo Tabular, an open-source model that predicts tabular outcomes without training. It ranks first on four benchmarks and runs 17 times faster than competitor LimiX-2.
Read more →
Ant Group introduces Ling-3.1-flash with a one-million-token context window
Ant Group announced Ling-3.1-flash, a mixture-of-experts AI model with a one-million-token context window designed for long-document processing in education, healthcare, and research, with an open release planned.
Read more →
Startup founder who lost his voice builds wearable device that restores emotional speech
Konrad Zieliński built a $500-$1,500 wearable that restores emotional speech for laryngectomy patients without surgery. His startup, Uhura Bionics, will demo it at TechCrunch Disrupt 2026.
Read more →
Latest AI News for Hospitality and Events
Inception launches Mercury Voice for low-latency agent conversations
Inception launched Mercury Voice, a low-latency conversational AI using diffusion modeling. The report advises customer support and hospitality teams to test it under real traffic, focusing on interruption handling, accent recognition, safety, and escalation.
Read more →
DoorDash launches AI agent for text-based food ordering
DoorDash launched a text-to-order AI agent integrating with Apple Messages to compete with Uber Eats and Grubhub. The company also announced drone delivery testing in Northern California.
Read more →
fal adds Pixelcut looping video for product animations
fal’s Pixelcut looping video model converts static product images into short, repeating animations with a fixed camera angle, enabling teams to generate dynamic marketing assets while requiring manual quality checks and disclosure of AI use.
Read more →
Instinct rolls out product recommendations amid user backlash
Instinct Selections pushes unsolicited product recommendations to users of the $10 billion AI startup. The feature, launched after a $1 billion Series C, has triggered backlash from investors who compare the spam-like alerts to intrusive advertising.
Read more →
Airbnb adds on-demand services and AI search to travel app
Airbnb launched AI-powered search and expanded on-the-ground services to compete with hotels. CEO Brian Chesky cited the challenge of maintaining conversion rates across a $100 billion platform.
Read more →
HeyGen launches a low-cost prompt-to-video API for business content
HeyGen launched a low-cost API that generates synchronized video and audio from text or images, allowing teams to create content without traditional filming, though human review for brand and legal compliance is still required.
Read more →
Latest AI News for Human Resources
Destro raises $8 million to deploy AI logistics layer instead of building robots
Destro raised $8 million to deploy software that orchestrates existing robots and humans in logistics. The startup is expanding from a three-robot pilot to 26 robots at a Yusen Logistics facility.
Read more →
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Speech to Text API adds experimental transcript editing via natural-language instructions
ElevenLabs added transcript editing to its API, applying natural-language commands to clean data in one request. This feature costs 30% more than standard transcription.
Read more →
Meta denies Muse read private messages without consent
Meta denies its Muse AI accessed private messages without consent, claiming integration requires three separate permission steps. The dispute follows a New Mexico jury verdict finding the company misled users about data practices in the Cambridge Analytica scandal.
Read more →
Consumer AI makes a comeback but faces hard economic limits
Only 2.2% of consumers pay for AI, averaging $31 monthly. Even with 325 million subscribers, revenue would remain below a third of OpenAI's operating costs.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
Latest AI News for Insurance
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Meta denies Muse read private messages without consent
Meta denies its Muse AI accessed private messages without consent, claiming integration requires three separate permission steps. The dispute follows a New Mexico jury verdict finding the company misled users about data practices in the Cambridge Analytica scandal.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
NVIDIA releases Kumo Tabular, an open foundation model for tabular data that tops four benchmarks
NVIDIA released Kumo Tabular, an open-source model that predicts tabular outcomes without training. It ranks first on four benchmarks and runs 17 times faster than competitor LimiX-2.
Read more →
Latest AI News for IT and Development
Cloudflare rebuilds Containers for faster persistent agent sandboxes
Cloudflare rebuilt its Containers platform to speed up agent startup times by six times and added runtime configuration and filesystem snapshots to better support persistent, stateful workloads.
Read more →
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
Latest AI News for Legal
Trump shares AI pledge with "United States" misspelled as "Unites States"
Trump and AI execs signed the non-binding "Joint Commitment on Frontier Responsibilities." The accompanying White House photo featured a typo, misspelling the country as "Unites States."
Read more →
Meta denies Muse read private messages without consent
Meta denies its Muse AI accessed private messages without consent, claiming integration requires three separate permission steps. The dispute follows a New Mexico jury verdict finding the company misled users about data practices in the Cambridge Analytica scandal.
Read more →
Instinct rolls out product recommendations amid user backlash
Instinct Selections pushes unsolicited product recommendations to users of the $10 billion AI startup. The feature, launched after a $1 billion Series C, has triggered backlash from investors who compare the spam-like alerts to intrusive advertising.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
Latest AI News for Management
Destro raises $8 million to deploy AI logistics layer instead of building robots
Destro raised $8 million to deploy software that orchestrates existing robots and humans in logistics. The startup is expanding from a three-robot pilot to 26 robots at a Yusen Logistics facility.
Read more →
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Flow Engineering raises $50M at a $750M valuation
Flow Engineering raised $50 million in a Series B round at a $750 million valuation. The capital will expand its AI agents for hardware design, already used by Anduril and Rivian.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
Consumer AI makes a comeback but faces hard economic limits
Only 2.2% of consumers pay for AI, averaging $31 monthly. Even with 325 million subscribers, revenue would remain below a third of OpenAI's operating costs.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
Latest AI News for Marketing
fal adds Pixelcut looping video for product animations
fal’s Pixelcut looping video model converts static product images into short, repeating animations with a fixed camera angle, enabling teams to generate dynamic marketing assets while requiring manual quality checks and disclosure of AI use.
Read more →
Instinct rolls out product recommendations amid user backlash
Instinct Selections pushes unsolicited product recommendations to users of the $10 billion AI startup. The feature, launched after a $1 billion Series C, has triggered backlash from investors who compare the spam-like alerts to intrusive advertising.
Read more →
Airbnb adds on-demand services and AI search to travel app
Airbnb launched AI-powered search and expanded on-the-ground services to compete with hotels. CEO Brian Chesky cited the challenge of maintaining conversion rates across a $100 billion platform.
Read more →
ElevenLabs valuation reaches $22 billion on enterprise demand
ElevenLabs doubled its valuation to $22 billion via a $300 million tender offer. Enterprise customers now drive 55% of revenue, handling over 15 million weekly conversations.
Read more →
Instagram adds AI assistant to Edits video editor
Instagram launched an AI assistant for its Edits app to analyze creator metrics and trends. Usage is capped unless users subscribe to Meta One for unlimited access.
Read more →
HeyGen launches a low-cost prompt-to-video API for business content
HeyGen launched a low-cost API that generates synchronized video and audio from text or images, allowing teams to create content without traditional filming, though human review for brand and legal compliance is still required.
Read more →
Latest AI News for Operations
Destro raises $8 million to deploy AI logistics layer instead of building robots
Destro raised $8 million to deploy software that orchestrates existing robots and humans in logistics. The startup is expanding from a three-robot pilot to 26 robots at a Yusen Logistics facility.
Read more →
Cloudflare rebuilds Containers for faster persistent agent sandboxes
Cloudflare rebuilt its Containers platform to speed up agent startup times by six times and added runtime configuration and filesystem snapshots to better support persistent, stateful workloads.
Read more →
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Runway introduces Praxis-1, an open-weight world action model for robotics
Runway released details of Praxis-1, an open-weight robotics model that uses video to solve data scarcity issues. The company reports a 0.95 correlation between simulated and real-world results, challenging traditional training methods.
Read more →
DoorDash launches AI agent for text-based food ordering
DoorDash launched a text-to-order AI agent integrating with Apple Messages to compete with Uber Eats and Grubhub. The company also announced drone delivery testing in Northern California.
Read more →
Flow Engineering raises $50M at a $750M valuation
Flow Engineering raised $50 million in a Series B round at a $750 million valuation. The capital will expand its AI agents for hardware design, already used by Anduril and Rivian.
Read more →
Latest AI News for PR and Communications
Speech to Text API adds experimental transcript editing via natural-language instructions
ElevenLabs added transcript editing to its API, applying natural-language commands to clean data in one request. This feature costs 30% more than standard transcription.
Read more →
Trump shares AI pledge with "United States" misspelled as "Unites States"
Trump and AI execs signed the non-binding "Joint Commitment on Frontier Responsibilities." The accompanying White House photo featured a typo, misspelling the country as "Unites States."
Read more →
Meta denies Muse read private messages without consent
Meta denies its Muse AI accessed private messages without consent, claiming integration requires three separate permission steps. The dispute follows a New Mexico jury verdict finding the company misled users about data practices in the Cambridge Analytica scandal.
Read more →
Consumer AI makes a comeback but faces hard economic limits
Only 2.2% of consumers pay for AI, averaging $31 monthly. Even with 325 million subscribers, revenue would remain below a third of OpenAI's operating costs.
Read more →
Instagram adds AI assistant to Edits video editor
Instagram launched an AI assistant for its Edits app to analyze creator metrics and trends. Usage is capped unless users subscribe to Meta One for unlimited access.
Read more →
America.gov chatbot recites Minecraft end poem when prompted
The U.S. launched its first public-facing AI chatbot, America.gov, built with Google and SpaceXAI. The bot resists jailbreaks, states Biden won 2020, and recites a 1,800-word government-themed rewrite of Minecraft's End Poem.
Read more →
Latest AI News for Product Development
Ideogram 4.5 reduces pixel drift in multi-turn edits
Ideogram 4.5 launched an image editing model that minimizes pixel drift during multi-turn sessions. The company claims it outperforms competitors that become unusable within a few edits.
Read more →
Restate secures $20 million in Series A funding to challenge Temporal in durable workflow infrastructure
Berlin startup Restate raised a $20 million Series A to expand its durable workflow engine for AI agents. The funding supports expansion against leader Temporal, which recently secured $550 million.
Read more →
Flow Engineering raises $50M at a $750M valuation
Flow Engineering raised $50 million in a Series B round at a $750 million valuation. The capital will expand its AI agents for hardware design, already used by Anduril and Rivian.
Read more →
Google launches Gemini 4 Argon for phased rollout to cyber defenders
Google unveiled Gemini 4 Argon at $2 per million input tokens, featuring a 1-million-token output limit. Early tests freed over 300 TiB of data center memory and replaced 32,000 lines of code.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
Latest AI News for Real Estate and Construction
Destro raises $8 million to deploy AI logistics layer instead of building robots
Destro raised $8 million to deploy software that orchestrates existing robots and humans in logistics. The startup is expanding from a three-robot pilot to 26 robots at a Yusen Logistics facility.
Read more →
Runway introduces Praxis-1, an open-weight world action model for robotics
Runway released details of Praxis-1, an open-weight robotics model that uses video to solve data scarcity issues. The company reports a 0.95 correlation between simulated and real-world results, challenging traditional training methods.
Read more →
Flow Engineering raises $50M at a $750M valuation
Flow Engineering raised $50 million in a Series B round at a $750 million valuation. The capital will expand its AI agents for hardware design, already used by Anduril and Rivian.
Read more →
Airbnb adds on-demand services and AI search to travel app
Airbnb launched AI-powered search and expanded on-the-ground services to compete with hotels. CEO Brian Chesky cited the challenge of maintaining conversion rates across a $100 billion platform.
Read more →
NVIDIA releases Kumo Tabular, an open foundation model for tabular data that tops four benchmarks
NVIDIA released Kumo Tabular, an open-source model that predicts tabular outcomes without training. It ranks first on four benchmarks and runs 17 times faster than competitor LimiX-2.
Read more →
HeyGen launches a low-cost prompt-to-video API for business content
HeyGen launched a low-cost API that generates synchronized video and audio from text or images, allowing teams to create content without traditional filming, though human review for brand and legal compliance is still required.
Read more →
Latest AI News for Sales
Inception launches Mercury Voice for low-latency agent conversations
Inception launched Mercury Voice, a low-latency conversational AI using diffusion modeling. The report advises customer support and hospitality teams to test it under real traffic, focusing on interruption handling, accent recognition, safety, and escalation.
Read more →
DoorDash launches AI agent for text-based food ordering
DoorDash launched a text-to-order AI agent integrating with Apple Messages to compete with Uber Eats and Grubhub. The company also announced drone delivery testing in Northern California.
Read more →
fal adds Pixelcut looping video for product animations
fal’s Pixelcut looping video model converts static product images into short, repeating animations with a fixed camera angle, enabling teams to generate dynamic marketing assets while requiring manual quality checks and disclosure of AI use.
Read more →
Instinct rolls out product recommendations amid user backlash
Instinct Selections pushes unsolicited product recommendations to users of the $10 billion AI startup. The feature, launched after a $1 billion Series C, has triggered backlash from investors who compare the spam-like alerts to intrusive advertising.
Read more →
Airbnb adds on-demand services and AI search to travel app
Airbnb launched AI-powered search and expanded on-the-ground services to compete with hotels. CEO Brian Chesky cited the challenge of maintaining conversion rates across a $100 billion platform.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
Latest AI News for Science and Research
Runway introduces Praxis-1, an open-weight world action model for robotics
Runway released details of Praxis-1, an open-weight robotics model that uses video to solve data scarcity issues. The company reports a 0.95 correlation between simulated and real-world results, challenging traditional training methods.
Read more →
Google launches Gemini 4 Argon for phased rollout to cyber defenders
Google unveiled Gemini 4 Argon at $2 per million input tokens, featuring a 1-million-token output limit. Early tests freed over 300 TiB of data center memory and replaced 32,000 lines of code.
Read more →
Fudan and Tencent launch ExplorationBench to test whether large language models can learn new rules through interaction
Fudan and Tencent released ExplorationBench, a benchmark testing AI rule-learning via interaction. Results show models reached 89.0% accuracy in code after exploration, proving feedback drives learning more than reasoning alone.
Read more →
NVIDIA releases Kumo Tabular, an open foundation model for tabular data that tops four benchmarks
NVIDIA released Kumo Tabular, an open-source model that predicts tabular outcomes without training. It ranks first on four benchmarks and runs 17 times faster than competitor LimiX-2.
Read more →
Ant Group introduces Ling-3.1-flash with a one-million-token context window
Ant Group announced Ling-3.1-flash, a mixture-of-experts AI model with a one-million-token context window designed for long-document processing in education, healthcare, and research, with an open release planned.
Read more →
Latest AI News for Writers
Ideogram 4.5 reduces pixel drift in multi-turn edits
Ideogram 4.5 launched an image editing model that minimizes pixel drift during multi-turn sessions. The company claims it outperforms competitors that become unusable within a few edits.
Read more →
Speech to Text API adds experimental transcript editing via natural-language instructions
ElevenLabs added transcript editing to its API, applying natural-language commands to clean data in one request. This feature costs 30% more than standard transcription.
Read more →
Google launches Gemini 4 Argon for phased rollout to cyber defenders
Google unveiled Gemini 4 Argon at $2 per million input tokens, featuring a 1-million-token output limit. Early tests freed over 300 TiB of data center memory and replaced 32,000 lines of code.
Read more →
Cohere launches Embed 5 and a retrieval metric focused on rank consistency
Cohere released Embed 5, featuring pro and fast tiers in a shared space, and introduced a new enterprise retrieval metric focused on rank consistency to better evaluate search performance.
Read more →
DeepSeek Harness preview adds desktop apps, plugins and scheduled automation
DeepSeek released a desktop preview of its Harness coding agent, adding plugin management and task automation while emphasizing enterprise needs like security isolation and audit trails.
Read more →
HeyGen launches a low-cost prompt-to-video API for business content
HeyGen launched a low-cost API that generates synchronized video and audio from text or images, allowing teams to create content without traditional filming, though human review for brand and legal compliance is still required.
Read more →