OpenAI dots
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 18
- WIRED Tests Six Always-On AI Agents; OpenAI Dots Wins Overall
WIRED’s software reporter spent 24 hours testing six always-on AI agents across task following, browser control, proactive behavior, pricing, and usability. OpenAI Dots ranked as the best overall, while Meta Muse placed second and was described as the easiest to set up; Anthropic’s Claude Duet received the best interface rating. The results are one reporter’s hands-on observations rather than a controlled benchmark, and the agents still made mistakes.
WIRED AI · 🔥 0 - Claude Duet joins the cloud-resident agent race
TMTPost reports that Anthropic launched Claude Duet on October 22 as a personal and work agent application, joining Meta Muse, OpenAI Dots, xAI Grok Bot and Manus Cue in giving AI cloud-hosted computers. In Anthropic's public demo, Duet pulled data from Shopify, Gusto, banking and other systems to produce a morning business briefing; the report presents this as a vendor demonstration rather than independently verified performance. The article argues that persistent cloud agents are shifting infrastructure demand toward CPU-heavy sandboxes, storage and higher-capacity networking, while noting that these benefits depend on sustainable agent economics.
钛媒体 · 🔥 0 - Personal AI agents face product, cost and ecosystem tests
A Best Ideas community compilation published by Overseas Unicorn examines personal AI agents, including Meta Muse, OpenAI dots, Instinct and Anthropic's Claude Duet. It argues that proactive task execution, cloud hosting and easier setup mark progress over chatbots, while the field remains early and lacks an “iPhone moment” for sustained mainstream use. The compilation also identifies high per-user service costs, platform restrictions, account authorization, permissions and privacy as unresolved commercial and ecosystem constraints. Its product, adoption and cost claims are community or practitioner views and estimates rather than independently verified findings.
虎嗅 AI · 🔥 0 - Muse growth fuels attention to personal AI agents
Meta’s Muse reached 10.9 million cumulative downloads after its September launch, including 10.2 million in the United States, according to Sensor Tower. Meta said about 25% of users use Muse daily, more than 25% shop through it, and the average order exceeds $200; these company-reported figures were not independently verified. Sensor Tower data showed day-14 retention falling to 22.5%, though still above the comparable-app average of under 20%. The reports also describe restrictions on access to third-party platforms, uncertainty over business models and user retention, and a macOS vulnerability. Meta patched the flaw and said it was a local privilege-escalation issue requiring malicious code already running on the device, making its risk lower than the researcher suggested.
虎嗅 AI · 🔥 0 - US Banks Restrict Third-Party AI Agents Accessing Customer Accounts
Major US banks including Morgan Stanley, JPMorgan, Bank of America, Wells Fargo and Citigroup are restricting third-party AI agents that access customer accounts with user-provided credentials, according to reports citing Axios. Measures include blocking some agents or limiting them to read-only access, while banks struggle to distinguish authorized agent activity from autonomous or unauthorized actions. Morgan Stanley said it has not observed any rogue AI agents on its platforms, while reports of agents acting beyond user instructions remain based partly on unnamed bank sources.
IT之家 AI · 🔥 0 - Joint survey defines a 6A framework for proactive AI agents
Researchers from UIUC, Google, Stanford, UC Berkeley and CMU presented the survey “Proactive AI for Superintelligent Agents,” which organizes proactive behavior around Awareness, Anticipation, Agenda, Arbitration, Action and Adaptation. 新智元 reported that the survey reviews evaluation methods and applications across 13 domains, while 36氪 人工智能 published an authorized syndication of the same article. The reports also cite Meta Muse, OpenAI dots, Proactive Agent and ProAgent as ongoing explorations of assistants that initiate useful work without waiting for a new instruction.
36氪 人工智能 · 🔥 0 - Meta's Muse and rival AI agents hit platform blockades as sites shut them out
Meta's Muse, xAI's Grok, OpenAI's Dots and startup Instinct's agent have been restricted or blocked on sites including Amazon, Yelp, eBay, Delta Air Lines, Adidas and JPMorgan Chase, according to a 智东西 report citing foreign media such as The Verge and TechCrunch. In China, Doubao's mobile assistant said 155 apps — among them WeChat, Taobao, Alipay, Meituan, Kimi and DeepSeek — refused its automated operations when its screen-automation declaration protocol (SAEP) closed public comment on October 14. Meta said refusing personal agents means refusing their customers and that it prefers working with retailers such as Walmart, Gap, Sephora and Best Buy, while Alibaba's Qwen Go keeps tasks inside its own Taobao and Alipay ecosystem; Meta and Sierra are also reported to be leading a personal-agent protocol that OpenAI and Anthropic have not joined.
智东西 · 🔥 0 - Dots Phone Is an OpenAI Dots Phone-Calling Plugin
A report identifies Dots Phone as a plugin for OpenAI Dots that supports phone calling. It provides no details about supported calling actions, regions, pricing, availability, technical requirements, or implementation.
Product Hunt · 🔥 0 - HKUST open-sources The Fourth Institute, an agent-society automated research platform
According to Xinzhiyuan, a team led by HKUST professors Zheng Guangting and Ma Songchen has open-sourced on GitHub an automated research platform called The Fourth Institute, where human researchers collaborate with AI agents through roles such as Resident, Lineage and Family, coordinated by the HATG ISA multi-agent specification and execution traces. The team says the platform has accumulated over 1.86 million thought and execution fragments, more than 930,000 interaction events, and over 5,100 hours of agent research and development work, and that it will progressively open-source those traces. As its first real project, the team applied the platform to embodied-AI infrastructure and says it optimized inference performance for dozens of models across multiple GPU platforms in 24 hours, with some results reaching SOTA; an Agentic OS for physical intelligence is also planned to be open-sourced.
新智元 · 🔥 0 - Manus Cue hands-on: agents get an identity, not yet always-on
Cue, the personal AI assistant Manus launched alongside Manus 2.0 on September 28, gives each agent its own email address, phone number, digital wallet and dedicated computer, and keeps working after the user logs out. After a week on a Cue Pro membership, Qidian Research Lab reported that email handling was the smoothest part, while phone and payment both required international accounts and could not be completed in its mainland China test environment, and scheduling could not reach WeChat, QQ or other everyday feeds. Multi-agent collaboration needed overseas app interfaces and a stable connection, group chats were unclear about who was answering or finishing, and on the same Q4 communications-plan prompt Cue's output was not notably better than Doubao or Tencent WorkBuddy, though the testers said the comparison was not fully equivalent. Cue is released for overseas users, a mainland version is still in preparation, and Cue membership is not interchangeable with Manus 2.0.
钛媒体 · 🔥 0 - Grok Bots appear on Grok web and mobile, users report
X user testingcatalog said Grok Bots have been rolled out on Grok web and mobile, calling it nice to have everything in one place. A reply said the bots are also being added to xchat inside the X app and web, while noting the surfaces lack feature parity; another user asked whether the desktop app's bots are the same set. The post also questioned whether xAI will keep both mobile apps long term, and none of these observations has been officially confirmed.
testingcatalog (X) · 🔥 0 - a16z's 'Wallet Wars': who lets AI agents pay as you
a16z published a newsletter by Terry Angelos titled "Wallet Wars" arguing that personal agents such as Instinct, OpenAI's Dots and Meta's Muse are moving assistants into text messages and familiar apps, and that completing tasks requires something intelligence alone cannot provide: permission to pay. The piece splits the contest in two: who helps an agent pay, and the more consequential question of who lets it transact as the user. It describes Stripe's Link handing agents single-use virtual cards, which can complete a purchase while breaking the link to the original card's loyalty benefits and merchant relationship, and points to card-network efforts — Visa Intelligent Commerce, Mastercard AgentPay and AMEX ACE — plus a proposed scoped agentic token held by a trusted wallet as a more open path. A Chinese translation was published by Huxiu AI, crediting the a16z original.
a16z News · 🔥 0 - Ryanair deploys predictive AI models to head off flight disruptions
CNET reports that Ryanair, Europe's largest airline, has implemented proprietary predictive AI models developed by its Ryanair Labs unit to anticipate when weather delays, airport curfews and crew-hours issues are likely to disrupt passengers' journeys before they escalate into widespread cancellations or diversions. CEO Eddie Wilson told CNET the goal is not to replace human operators but to give staff advanced tools for managing operations across 96 bases and more than 3,800 flights a day; he said the model can reassign crew or flights when a crew is about to exceed its allowed hours. The capability and efficiency claims come from the company, and the report does not disclose model accuracy or independent evaluations.
CNET · 🔥 0 - Why Anthropic Is Sitting Out the Personal AI Agent Race
A GeekPark analysis, republished by Huxiu, 36Kr and Cyzone, examines why Anthropic has no consumer always-on personal agent to match OpenAI's dots and Meta's Muse. It notes that OpenAI introduced dots at its Sept. 29 DevDay, with each dot running on GPT-6 Astra with its own cloud computer and browser and continuing work after ChatGPT is closed, while Meta launched Muse on Sept. 8 inside WhatsApp; Apptopia estimated about 2.8 million installs across platforms in 12 days and a brief No. 1 spot on the U.S. App Store. The article says Anthropic instead offers Claude Tag, a shared Claude in Slack channels released June 23 with its own identity and memory, plus Claude Code and Cowork; it reports Anthropic's claim that 65% of its product team's code came from an internal version of Claude Tag and attributes Anthropic's restraint to the lack of a large consumer entry point, a revenue mix of roughly 80% API and enterprise customers, always-on agent costs and safety concerns, while noting Anthropic has not publicly said it avoids personal agents over emotional dependence.
极客公园 · 🔥 0 - JPMorgan blocks external AI agents like Meta's Muse from client accounts
On its October 13 third-quarter earnings call, JPMorgan CEO Jamie Dimon said the bank will not let any outside AI agent into the company to access client accounts and act on their behalf, calling it too risky for clients and for JPMorgan. A JPMorgan spokesperson confirmed the bank currently bars third parties from logging in with customer bank credentials to act for retail clients, while saying it supports client-authorized data sharing through open banking providers — meaning agents can read some data such as balances via connectors like Plaid but cannot move money or pay bills. Dimon also drew a line on a "transfer of liability" when customers authorize outside agents, and CFO Jeremy Barnum said the bank began a targeted pilot of its in-house SmartCash tool in recent weeks.
cnBeta · 🔥 0 - 23-year-old's 14-person AI agent startup Instinct hits $10B valuation
Citing Reuters, 36Kr reports that Instinct, the personal AI assistant startup founded by 23-year-old Noah Shinn, announced a $1 billion funding round on September 28 with investors including Sequoia Capital, Benchmark and Coatue, at a reported valuation of about $10 billion — roughly 4x its $25 billion... valuation of $2.5 billion in late August — with a team of just 14 people. Rather than a standalone app, users reach the assistant by text or phone call to book flights, buy household goods, cancel subscriptions or contact merchants, and a group-chat agent feature has been opened to some users. Shinn has said annualized transaction volume on the platform is approaching $1 billion, more than half travel-related, though he did not explain the calculation and the figure is self-reported. The Instinct Selections recommendation feature launched in late September drew complaints after suggesting products such as luggage without a shopping request, leaving monetization, privacy and reliability still unproven.
36氪 人工智能 · 🔥 0 - Google's Gemini Agent can route tasks to Claude as it chases the agent race
At its Gemini at Work event on October 8, Google Cloud CEO Thomas Kurian introduced Gemini Agent, which pulls Gmail, Calendar, Docs, Drive and Chat into a single work agent and can also create “Coworker Agents” with their own email, calendar and Drive accounts. According to the report, the agent can route tasks between Google's own Gemini models and Anthropic's Claude, with more models planned, and is currently in private preview for select customers. The piece cites Google's own charts showing Gemini 4 Argon at 57.4% on Terminal-Bench 4.0 and 55.0% on FrontierSWE v2 versus Claude Opus 5.5 at 66.4% and 62.3%, and notes Anthropic is cultivating developers through the Claude Agent SDK, Claude Code Mods and API credits.
36氪 人工智能 · 🔥 0 - Personal AI assistant startup Tab exits stealth at $300M valuation
TechCrunch reported that US personal AI assistant startup Tab emerged from stealth on October 6 at a $300 million valuation, with SV Angel, Valar Ventures and American Spirit among its investors; the company did not disclose the actual amount raised, nor its user numbers, revenue or paid usage. Tab is operated by Terrasoft, a Y Combinator Winter 2020 batch company, and users can send instructions via iMessage or WhatsApp to order food, plan trips, buy gifts, make calls or pay bills; it has its own phone number, computer and wallet, and can connect accounts including Gmail, Google Calendar, Slack, GitHub, Linear and Notion, though it remains in closed beta. In a 15-task test by the independent Assistant Benchmark site, Tab received 97 messages over two days with a median reply time of 12 seconds, scoring 5.9 overall and ranking 21st: it earned 10 for a phone task and 9 for permissions and privacy, but only 3 for restaurant booking by web and phone, 5 for a travel task and 1 for a purchase task, where card payments failed repeatedly. The leaderboard's top three were Meta Muse at 9.3, Instinct at 8.5 and OpenAI Dots at 8.4.
36氪 人工智能 · 🔥 0
Experience and discussion from the community
Share my OpenAI dots experienceAsk about OpenAI dots
Nobody has shared their experience with OpenAI dots yet.