[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"post-chatgpt-voice-mode-makes-conversations-feel-genuinely-human":3,"nav-categories":49,"trending-posts":74,"comments-177":210,"related-177":211},{"id":4,"slug":5,"title":6,"excerpt":7,"content":8,"date":9,"modifiedDate":10,"authorSlug":11,"featuredImage":12,"thumbnailImage":13,"smallImage":14,"formattedDate":15,"readingTime":16,"author":17,"categories":21,"tags":30,"seo":33},177,"chatgpt-voice-mode-makes-conversations-feel-genuinely-human","ChatGPT Voice Mode makes conversations feel genuinely human","Making voice feel less like an operating software and more like talking to someone.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">OpenAI is rolling out GPT-Live that powers \u003Ca href=\"https://www.neuralbriefly.com/category/chatgpt\" target=\"_blank\" rel=\"noopener\">ChatGPT\u003C/a> Voice with more natural, real-time conversations. Unlike older versions that made you wait your turn to speak, GPT-Live listens and responds at the same time, so you can interrupt, pause mid-thought, or stay quiet without triggering an awkward response. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">If you use ChatGPT Voice to brainstorm, practice a language, or get hands-free answers, this update should feel noticeably smoother for your very next conversation. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">Talking to ChatGPT finally feels normal\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The update fixes two common complaints about AI voice chats: slow replies and rigid back-and-forth. It means conversations with ChatGPT should feel much closer to speaking with another person than just issuing commands to a voice assistant.\u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>ALSO READ: \u003Ca href=\"https://www.neuralbriefly.com/blog/chatgpts-scheduled-tasks-makes-ai-do-the-work-while-you-sleep\">ChatGPT’s Scheduled Tasks makes AI do the work while you sleep\u003C/a>\u003C/strong>\u003C/p>\n\u003Cdiv class=\"flex items-center justify-between gap-4 mb-4 flex-wrap\" data-v-ae809a22=\"\">\n\u003Cdiv class=\"flex items-center gap-2 text-[12px] text-gray-500 dark:text-gray-400\" data-v-ae809a22=\"\">This is starkly different to previous version—where you speak, the AI processes, and then responds. GPT-Live’s full-duplex architecture enables the model to listen what you’re saying.\u003C/div>\n\u003C/div>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The updated ChatGPT Voice  unlocks several improvements and practical changes\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">– Receive natural acknowledgements like mm-hmm while speaking\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">– Handles background noise better\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">– No more question-answer exchanges\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Most users should already have acess to the updated ChatGPT Voice mode. Paid users get GPT-Live-1 whereas free subscribers get a lighter version, GPT-Live-1 mini. You can trigger the feature by just tapping the voice button, no settings are required. \u003C/span>\u003C/p>\n\u003Cp>\u003Cb>Why this matters: \u003C/b>\u003Cspan style=\"font-weight: 400;\">Conversational AI is gaining traction lately to maintain natural dialogue instead of simply answering questions.\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Source: \u003C/span>\u003Ca href=\"https://openai.com/index/introducing-gpt-live/\" target=\"_blank\" rel=\"noopener\">\u003Cspan style=\"font-weight: 400;\">OpenAI\u003C/span>\u003C/a>\u003C/p>\n\u003Cdiv class=\"flex items-center justify-between gap-4 mb-4 flex-wrap\" data-v-ae809a22=\"\">\u003C/div>","2026-07-10T20:55:28","2026-07-10T21:01:25","ketangopal","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-10-2026-at-08_49_43-PM-e1783697002454.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-10-2026-at-08_49_43-PM-768x512.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-10-2026-at-08_49_43-PM-300x200.png","July 10, 2026",2,{"id":16,"name":18,"avatarUrl":19,"description":20},"Ketan Gopal","https://secure.gravatar.com/avatar/f6854931dad97567411fa56dcdfe969cdbf8d1acb46b18669260a0936f07dd4d?s=96&d=mm&r=g","Ketan is a Full Stack Developer with 8 years of experience in the tech industry. He is obsessed with the rapidly evolving world of artificial intelligence and always looking to test out new tools and features.",[22,26],{"id":23,"name":24,"slug":25,"count":-1},25,"ChatGPT","chatgpt",{"id":27,"name":28,"slug":29,"count":-1},4,"News","news",[31],{"id":32,"name":24,"slug":25},21,{"title":34,"description":35,"ogImage":12,"ogImageWidth":36,"ogImageHeight":37,"ogType":38,"breadcrumbs":39,"article":47},"ChatGPT Voice mode makes AI conversations less robotic","OpenAI has launched GPT-Live, bringing more natural voice conversations to ChatGPT with real-time responses, interruptions, and lower latency.",1200,800,"article",[40,43,45],{"name":41,"url":42},"Home","/",{"name":24,"url":44},"/category/chatgpt",{"name":6,"url":46},"/blog/chatgpt-voice-mode-makes-conversations-feel-genuinely-human",{"publishedTime":9,"modifiedTime":10,"author":18,"tags":48},[24],[50,52,56,60,64,68,72],{"id":23,"name":24,"slug":25,"count":51},5,{"id":53,"name":54,"slug":55,"count":51},26,"Claude","claude",{"id":51,"name":57,"slug":58,"count":59},"Features","features",1,{"id":61,"name":62,"slug":63,"count":27},27,"Gemini","gemini",{"id":65,"name":66,"slug":67,"count":59},6,"How-to","how-to",{"id":69,"name":70,"slug":71,"count":59},9,"Latest","latest",{"id":27,"name":28,"slug":29,"count":73},12,[75,89,122,152,182],{"id":4,"slug":5,"title":6,"excerpt":7,"content":8,"date":9,"modifiedDate":10,"authorSlug":11,"featuredImage":12,"thumbnailImage":13,"smallImage":14,"formattedDate":15,"readingTime":16,"author":76,"categories":77,"tags":80,"seo":82},{"id":16,"name":18,"avatarUrl":19,"description":20},[78,79],{"id":23,"name":24,"slug":25,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[81],{"id":32,"name":24,"slug":25},{"title":34,"description":35,"ogImage":12,"ogImageWidth":36,"ogImageHeight":37,"ogType":38,"breadcrumbs":83,"article":87},[84,85,86],{"name":41,"url":42},{"name":24,"url":44},{"name":6,"url":46},{"publishedTime":9,"modifiedTime":10,"author":18,"tags":88},[24],{"id":90,"slug":91,"title":92,"excerpt":93,"content":94,"date":95,"modifiedDate":95,"authorSlug":96,"featuredImage":97,"thumbnailImage":98,"smallImage":99,"formattedDate":100,"readingTime":16,"author":101,"categories":105,"tags":108,"seo":111},174,"claude-cowork-is-no-longer-tied-to-your-desktop","Claude Cowork is no longer tied to your desktop","Anthropic just untethered its AI task-delegation tool from the desktop. Your assigned work now survives a closed laptop lid.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">Anthropic has announced that \u003Ca href=\"https://www.neuralbriefly.com/category/claude\">Claude\u003C/a> Cowork will run natively on web and mobile. It will allow users to start long-running tasks on one device and continue monitoring them from another. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">Cowork works on mobile and web now\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Until now, Cowork only worked when your desktop app stayed open, which means closing the laptop paused whatever Claude was doing. That limitation is gone. Tasks can now run with no device online at all, so you could set a Monday morning briefing for 6 a.m. and find it finished, sourced from emails, transcripts, and news, with a follow-up draft waiting for your approval over coffee.\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">When Claude hits a decision only you can make, it pings your phone and waits. Nothing ships without your sign-off, whether you’re redirecting a draft mid-meeting or reviewing final output later.\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">\u003Cstrong>The bigger shift:\u003C/strong> One interesting insight Anthropic shared is that more than 90% of Cowork sessions aren’t for software development. It was for ordinary knowledge like business operations, content creation, contract reviews, expense reconciliation, and client decks. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">This means that Anthropic is framing Cowork as a general-purpose AI colleague that can handle multi-step office work with minimal supervision.\u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>ALSO READ: \u003Ca href=\"https://www.neuralbriefly.com/blog/claude-sonnet-5-offers-near-opus-performance-at-a-fraction-of-the-cost\">Claude Sonnet 5 offers near-Opus performance at a fraction of the cost\u003C/a>\u003C/strong>\u003C/p>\n\u003Cdiv class=\"flex items-center justify-between gap-4 mb-4 flex-wrap\" data-v-ae809a22=\"\">\n\u003Cdiv class=\"flex items-center gap-2 text-[12px] text-gray-500 dark:text-gray-400\" data-v-ae809a22=\"\">The desktop version still offers the fullest set of features like access to local files and browser-based workflows.\u003C/div>\n\u003C/div>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Access to Cowork is rolling out to Max subscribers with support for more plans over the coming weeks. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">By untethering Cowork from the desktop, the company is betting that the next generation of AI assistants won’t just answer questions—they’ll keep working in the background until the job is done.\u003C/span>\u003C/p>\n\u003Cp>\u003Ca href=\"https://claude.com/blog/cowork-web-mobile\">Source\u003C/a>\u003C/p>","2026-07-09T16:19:00","g-viiirajgmail-com","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-9-2026-at-04_18_13-PM-e1783594129324.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-9-2026-at-04_18_13-PM-768x512.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-9-2026-at-04_18_13-PM-300x200.png","July 9, 2026",{"id":59,"name":102,"avatarUrl":103,"description":104},"Viraj Gawde","https://secure.gravatar.com/avatar/f5b758924053471549105f52fd998f20607c74cea7d5f826acd8b57d4480c1a7?s=96&d=mm&r=g","Viraj is a Writer, Editor, and Strategist at large. With over eight years of experience, he has bylines at reputable Indian and global publications. Viraj's expertise in tech spans across news coverage, how-to guides, and op-eds.",[106,107],{"id":53,"name":54,"slug":55,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[109],{"id":110,"name":54,"slug":55},19,{"title":112,"description":113,"ogImage":97,"ogImageWidth":36,"ogImageHeight":37,"ogType":38,"breadcrumbs":114,"article":120},"Claude Cowork now runs on mobile and browser","Claude Cowork no longer needs your laptop open. Tasks now run on web and mobile, and keep going while you're away.",[115,116,118],{"name":41,"url":42},{"name":54,"url":117},"/category/claude",{"name":92,"url":119},"/blog/claude-cowork-is-no-longer-tied-to-your-desktop",{"publishedTime":95,"modifiedTime":95,"author":102,"tags":121},[54],{"id":123,"slug":124,"title":125,"excerpt":126,"content":127,"date":128,"modifiedDate":129,"authorSlug":96,"featuredImage":130,"thumbnailImage":131,"smallImage":132,"formattedDate":133,"readingTime":16,"author":134,"categories":135,"tags":138,"seo":140},164,"claude-sonnet-5-offers-near-opus-performance-at-a-fraction-of-the-cost","Claude Sonnet 5 offers near-Opus performance at a fraction of the cost","Narrows the gap with Opus 4.8 on coding and agentic tasks—without the Opus price tag.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">Anthropic has released Claude Sonnet 5 positioning it as the company’s most agentic AI model to date. It closes the performance gap with Anthropic’s flagship Opus 4.8 model while staying well priced below it. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Sonnet 5’s launch matters for developers, power users, and businesses running demanding AI workloads. It is now the default model for Free and Pro users. It’s also live for Max, Team, and Enterprise plans, alongside \u003Ca href=\"https://www.neuralbriefly.com/category/claude\">Claude\u003C/a> Code and Claude Platform. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">A stronger Claude for coding, research, and everyday automation\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Claude Sonnet 5 doesn’t deliver just better answers. It is designed to take on work that requires reasoning, tools, and self-checking along the way. For developers, this means debugging a stubborn issue, writing and testing a code, or making multi-step changes across a project. For businesses, it could help with tasks such as analysing data, researching information, updating records, and handling routine workflows that previously needed repeated prompts\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Sonnet 5 is claimed to handle longer, messier tasks more reliably and is less likely to make things up compared to Sonnet 4.6 that often needed hand-holding or produced results that required double-checking. \u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>ALSO READ: \u003C/strong>\u003Ca href=\"https://www.neuralbriefly.com/blog/claude-tag-wants-to-become-your-slack-teammate-and-its-not-just-a-chatbot-window\">Claude Tag wants to become your Slack teammate and it’s not just a chatbot window\u003C/a>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">It is also harder to trick with malicious prompts hidden in documents or web pages—a real concern for anyone using AI agents that browse the web or read files on their behalf.\u003C/span>\u003C/p>\n\u003Cp>\u003Cb>The practical upshot:\u003C/b>\u003Cspan style=\"font-weight: 400;\"> tasks that previously required its larger and pricer Opus models to get trustworthy results can now often be handled by Sonnet 5 instead. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Sonnet 5 launches at an introductory pricing of $2 per million input tokens and $10 per million output tokens till August 31. Post that, it will be priced $3 and $15 respectively.  \u003C/span>\u003C/p>\n\u003Cp>\u003Cb>Bottom line: \u003C/b>\u003Cspan style=\"font-weight: 400;\">Sonnet 5 makes near-frontier agentic performance accessible without an Opus-level bill. Users should consider benchmarking it against their current Sonnet 4.6 workloads before the introductory pricing expires.\u003C/span>\u003C/p>","2026-07-01T19:53:16","2026-07-01T19:54:32","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-1-2026-at-07_52_10-PM-e1782915859433.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-1-2026-at-07_52_10-PM-768x512.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/07/ChatGPT-Image-Jul-1-2026-at-07_52_10-PM-300x200.png","July 1, 2026",{"id":59,"name":102,"avatarUrl":103,"description":104},[136,137],{"id":53,"name":54,"slug":55,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[139],{"id":110,"name":54,"slug":55},{"title":141,"description":142,"ogImage":130,"ogImageWidth":143,"ogImageHeight":144,"ogType":38,"breadcrumbs":145,"article":150},"Claude Sonnet 5 makes autonomous AI work affordable","Anthropic's Claude Sonnet 5 brings agentic capabiltiies with less hand-holding and without pushing users towards its most expensive model.",900,600,[146,147,148],{"name":41,"url":42},{"name":54,"url":117},{"name":125,"url":149},"/blog/claude-sonnet-5-offers-near-opus-performance-at-a-fraction-of-the-cost",{"publishedTime":128,"modifiedTime":129,"author":102,"tags":151},[54],{"id":153,"slug":154,"title":155,"excerpt":156,"content":157,"date":158,"modifiedDate":159,"authorSlug":96,"featuredImage":160,"thumbnailImage":161,"smallImage":162,"formattedDate":163,"readingTime":16,"author":164,"categories":165,"tags":168,"seo":171},158,"gemini-in-google-meet-now-writes-meeting-minutes-but-only-if-you-pay-up","Gemini in Google Meet now writes meeting minutes, but only if you pay up","Gemini types up the minutes of meeting and emails you the recap.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">Most meeting notes are either incomplete or force someone to stop paying attention just to keep up. Google is now tackling this issue by bringing \u003Ca href=\"https://www.neuralbriefly.com/category/gemini\">Gemini\u003C/a>’s ‘Take notes for me’ feature in Google Meet to individual users. For anyone who has scrambled to type notes while also trying to listen, the new feature closes that gap by handing the job entirely to AI. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Previously, it was limited to education, enterprise, and workspace users. The only feature is rolling out on web and mobile for Google AI Pro and Ultra subscribers. So, free Google AI users are out of luck. Another catch is that it is available in select languages for now. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">You can focus on your calls without worrying about meeting notes\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Gemini in Google Meet transcribes the conversations in real-time, identifies the key discussion points, and compiles it into a structured summary with clear actionable items. Notes are automatically saved to a Google Doc in the host’s Google Drive. Participants also receive a follow-up email containing the summary and action items once the meeting ends. They are notified whenever the feature is switched on for a call. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">During a live call, users can enable the ‘take notes for me’ feature by selecting the pencil icon at the top of the Meet window. It can also be permanently enabled through the Meeting records in Google Meet settings. \u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>ALSO READ: \u003Ca href=\"https://www.neuralbriefly.com/blog/gemini-3-5-flash-gains-computer-use-for-browser-and-desktop-automation\">Gemini 3.5 Flash gains Computer Use for browser and desktop automation\u003C/a>\u003C/strong>\u003C/p>\n\u003Cp>\u003Cstrong>Between the lines:\u003C/strong> This is a small but practical update for anyone running frequent calls—client meetings, planning sessions, team stand-ups, and any other meeting where decisions can get lost once it concludes.\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Rather than turning Meet into another recording tool, Gemini creates a usable follow-up document without leaving one person responsible for writing it all down. \u003C/span>\u003C/p>","2026-06-30T23:07:06","2026-07-01T20:12:48","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-30-2026-at-11_03_55-PM-e1782841062760.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-30-2026-at-11_03_55-PM-768x512.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-30-2026-at-11_03_55-PM-300x200.png","June 30, 2026",{"id":59,"name":102,"avatarUrl":103,"description":104},[166,167],{"id":61,"name":62,"slug":63,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[169],{"id":170,"name":62,"slug":63},29,{"title":172,"description":173,"ogImage":160,"ogImageWidth":143,"ogImageHeight":144,"ogType":38,"breadcrumbs":174,"article":180},"Gemini in Google Meet now writes meeting minutes","Gemini in Google Meet can now generate minutes of meeting for AI Pro and Ultra subscribers. It gets delivered to the host and participants.",[175,176,178],{"name":41,"url":42},{"name":62,"url":177},"/category/gemini",{"name":155,"url":179},"/blog/gemini-in-google-meet-now-writes-meeting-minutes-but-only-if-you-pay-up",{"publishedTime":158,"modifiedTime":159,"author":102,"tags":181},[62],{"id":183,"slug":184,"title":185,"excerpt":186,"content":187,"date":188,"modifiedDate":189,"authorSlug":11,"featuredImage":190,"thumbnailImage":191,"smallImage":192,"formattedDate":193,"readingTime":16,"author":194,"categories":195,"tags":198,"seo":200},154,"gemini-3-5-flash-gains-computer-use-for-browser-and-desktop-automation","Gemini 3.5 Flash gains Computer Use for browser and desktop automation","Google most capable fast model became more powerful. It allows developers to build AI agents that can operate certain tasks autonomously.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">Google has now built Computer Use in \u003Ca href=\"https://www.neuralbriefly.com/category/gemini\">Gemini\u003C/a> 3.5 Flash. Instead of a consumer-facing feature, it is aimed at developers who can create AI agents that can see a screen, reason about what’s on it, and take action—clicking, scrolling, and typing. This isn’t a separate product or an add-on. It’s baked directly into the model devs are already using. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">In practice, it shifts Gemini from simply suggesting the next step to helping software carry it out. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">Computer use in Gemini 3.5 Flash\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Previously limited to Gemini 2.5 Flash, developers can now build custom agents that operate software on your behalf across browser, mobile, and desktop environments. This means browsing the web, filling forms, testing apps, or navigating desktop tools without any custom integrations. The most practical use cases of computer use are enterprise-grade: continuous software testing and knowledge work across professional applications. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Developers and enterprise users can start using the computer used in 3.5 Flash via the Gemini API and Gemini Enterprise Agent platform.\u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>ALSO READ: \u003Ca href=\"https://www.neuralbriefly.com/blog/claude-tag-wants-to-become-your-slack-teammate-and-its-not-just-a-chatbot-window\">Claude Tag wants to become your Slack teammate and it’s not just a chatbot window\u003C/a>\u003C/strong>\u003C/p>\n\u003Cdiv class=\"flex items-center justify-between gap-4 mb-4 flex-wrap\" data-v-ae809a22=\"\">\n\u003Cdiv class=\"flex items-center gap-2 text-[12px] text-gray-500 dark:text-gray-400\" data-v-ae809a22=\"\">Google is using targeted adversarial training for computer use in Gemini 3.5 Flash to mitigate prompt injection risk for agents operating in live environments. Two optional enterprise safeguards let businesses require user confirmation before sensitive actions and automatically halt tasks if an injection attempt is detected.\u003C/div>\n\u003C/div>\n\u003Cp>Source: \u003Ca href=\"https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-computer-use-gemini-3-5-flash/\">Google\u003C/a>\u003C/p>","2026-06-27T14:46:33","2026-06-27T14:47:29","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-27-2026-at-02_42_27-PM-e1782551703469.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-27-2026-at-02_42_27-PM-768x512.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-27-2026-at-02_42_27-PM-300x200.png","June 27, 2026",{"id":16,"name":18,"avatarUrl":19,"description":20},[196,197],{"id":61,"name":62,"slug":63,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[199],{"id":170,"name":62,"slug":63},{"title":201,"description":202,"ogImage":190,"ogImageWidth":143,"ogImageHeight":144,"ogType":38,"breadcrumbs":203,"article":208},"Gemini 3.5 Flash gains Computer Use but it's only for developers","Gemini 3.5 Flash lets developers build AI agents that can operate browsers, mobile apps, and desktop software autonomously.",[204,205,206],{"name":41,"url":42},{"name":62,"url":177},{"name":185,"url":207},"/blog/gemini-3-5-flash-gains-computer-use-for-browser-and-desktop-automation",{"publishedTime":188,"modifiedTime":189,"author":18,"tags":209},[62],[],[212,240,271],{"id":213,"slug":214,"title":215,"excerpt":216,"content":217,"date":218,"modifiedDate":219,"authorSlug":11,"featuredImage":220,"thumbnailImage":221,"smallImage":222,"formattedDate":223,"readingTime":16,"author":224,"categories":225,"tags":228,"seo":230},125,"chatgpts-scheduled-tasks-makes-ai-do-the-work-while-you-sleep","ChatGPT’s Scheduled Tasks makes AI do the work while you sleep","Instead of opening ChatGPT every day to ask the same thing, users can now set tasks once and let the assistant handle the follow-up.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">Most people treat ChatGPT has a back-and-forth tool. You remember the task, type the same instruction again, and it answers. If you’re still manually checking for updates, writing the same recurring prompts, or relying on separate reminder apps, you’re leaving automation on the table. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">OpenAI is now fixing this with an upgraded Scheduled Tasks experience inside ChatGPT’s sidebar as a dedicated page. The update makes \u003Ca href=\"https://www.neuralbriefly.com/category/chatgpt\">ChatGPT\u003C/a> as an assistant that follows up for you. It’s actually hard to miss and signficantly more powerful. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">ChatGPT gets a proper task dashboard\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Previously, scheduled tasks existed but were buried. So, many users didn’t know ChatGPT could handle recurring prompts for all. The new update adds a visible shortcut in the sidebar. The dedicated Scheduled page gives users one place to view active tasks, check when they’ll run, along with broader options like pause, resume, edit, and delete. \u003C/span>\u003C/p>\n\u003Cp>\u003Cimg loading=\"lazy\" decoding=\"async\" class=\"aligncenter wp-image-128 size-full\" src=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks.jpg\" alt=\"ChatGPT Scheduled Tasks now stays in the sidebar\" width=\"2100\" height=\"1400\" srcset=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks.jpg 2100w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks-300x200.jpg 300w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks-1024x683.jpg 1024w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks-768x512.jpg 768w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks-1536x1024.jpg 1536w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Scheduled-Tasks-2048x1365.jpg 2048w\" sizes=\"auto, (max-width: 2100px) 100vw, 2100px\" />\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The end benfit for users is clear—you can ask ChatGPT to remind you to follow up with a client tomorrow or send a daily news briefing. But the biggest impact is with monitoring tasks that replaces the current Pulse system. It can monitor for changes and notify only if there’s a meaningful update like when a package ships or a price drops, instead of getting pinged every hour. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Tasks can be scheduled for specific times or broader windows—morning, afternoon, or evening. \u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>READ ALSO: \u003Ca href=\"https://www.neuralbriefly.com/blog/how-to-send-emails-directly-from-chatgpt-without-opening-gmail-or-outlook\">How to send emails directly from ChatGPT without opening Gmail or Outlook\u003C/a>\u003C/strong>\u003C/p>\n\u003Cp>OpenAI is rolling out the upgraded Scheduled Tasks experience to paid users. That includes Go, Plus, Pro, Business, and Enterprise. It is available on both mobile and web platforms. One caveat worth noting is that Tasks cannot run more than once per hour.\u003C/p>","2026-06-18T20:29:35","2026-06-18T23:38:00","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/chatgpt_scheduled_tasks_900x600.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/chatgpt_scheduled_tasks_900x600-768x512.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/chatgpt_scheduled_tasks_900x600-300x200.png","June 18, 2026",{"id":16,"name":18,"avatarUrl":19,"description":20},[226,227],{"id":23,"name":24,"slug":25,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[229],{"id":32,"name":24,"slug":25},{"title":231,"description":232,"ogImage":220,"ogImageWidth":143,"ogImageHeight":144,"ogType":38,"breadcrumbs":233,"article":238},"ChatGPT's Scheduled Tasks fixes AI's biggest annoyances","ChatGPT's Scheduled Tasks lets you automate reminders and recurring work. It puts the most underused feature in the sidebar.",[234,235,236],{"name":41,"url":42},{"name":24,"url":44},{"name":215,"url":237},"/blog/chatgpts-scheduled-tasks-makes-ai-do-the-work-while-you-sleep",{"publishedTime":218,"modifiedTime":219,"author":18,"tags":239},[24],{"id":241,"slug":242,"title":243,"excerpt":244,"content":245,"date":246,"modifiedDate":247,"authorSlug":96,"featuredImage":248,"thumbnailImage":249,"smallImage":250,"formattedDate":251,"readingTime":252,"author":253,"categories":254,"tags":257,"seo":259},112,"how-to-send-emails-directly-from-chatgpt-without-opening-gmail-or-outlook","How to send emails directly from ChatGPT without opening Gmail or Outlook","ChatGPT is already helpful for writing emails. You can ask it to draft a message, tweak the tone, and even generate a polished response. However, the main friction point is in copying everything, opening your email client, pasting the text, and then hitting send yourself. This workflow is finally changing, as OpenAI has introduced a […]","\u003Cp>\u003Cspan style=\"font-weight: 400;\">ChatGPT is already helpful for writing emails. You can ask it to draft a message,  tweak the tone, and even generate a polished response. However, the main friction point is in copying everything, opening your email client, pasting the text, and then hitting send yourself. This workflow is finally changing, as OpenAI has introduced a new feature that allows you to send emails directly from the web version of ChatGPT. This turns the chatbot into something closer to a true AI assistant.\u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">We tested out the new feature and can confirm that it indeed works smoothly. If you are eager to try it yourself, here’s a step-by-step process of how you can send emails directly from \u003Ca href=\"https://www.neuralbriefly.com/category/chatgpt\">ChatGPT\u003C/a>.\u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">How to send emails directly from ChatGPT?\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">It’s worth noting that your ChatGPT account should be connected to an email account. At the moment, it can only work with Gmail and Outlook accounts. Other email clients aren’t supported.\u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>Step 1: Connect your email account\u003C/strong>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Click your profile icon in the bottom left sidebar and open \u003C/span>\u003Cb>Settings\u003C/b>\u003Cspan style=\"font-weight: 400;\">. From there navigate to \u003C/span>\u003Cb>Apps\u003C/b>\u003Cspan style=\"font-weight: 400;\"> from the list and select \u003C/span>\u003Cb>Add More\u003C/b>\u003Cspan style=\"font-weight: 400;\"> option. Search for Gmail or Outlook and follow the prompts to connect your account. \u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>Step 2: Draft an email\u003C/strong>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Once your email is connected, ask ChatGPT to draft an email with a prompt. Do note that you are required to input the recipient’s mail address, either in the prompt or the chat response. You can also add CC and BCC recipients. \u003C/span>\u003C/p>\n\u003Cdiv id='gallery-1' class='gallery galleryid-112 gallery-columns-3 gallery-size-full'>\u003Cfigure class='gallery-item'>\n\t\t\t\u003Cdiv class='gallery-icon landscape'>\n\t\t\t\t\u003Ca href='https://lime-hamster-756747.hostingersite.com/?attachment_id=79'>\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"2524\" height=\"1280\" src=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM.png\" class=\"attachment-full size-full\" alt=\"Connect email account to ChatGPT\" srcset=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM.png 2524w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM-300x152.png 300w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM-1024x519.png 1024w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM-768x389.png 768w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM-1536x779.png 1536w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.27.26-PM-2048x1039.png 2048w\" sizes=\"auto, (max-width: 2524px) 100vw, 2524px\" />\u003C/a>\n\t\t\t\u003C/div>\u003C/figure>\u003Cfigure class='gallery-item'>\n\t\t\t\u003Cdiv class='gallery-icon landscape'>\n\t\t\t\t\u003Ca href='https://lime-hamster-756747.hostingersite.com/?attachment_id=80'>\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"1880\" height=\"1038\" src=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.34.42-PM.png\" class=\"attachment-full size-full\" alt=\"\" srcset=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.34.42-PM.png 1880w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.34.42-PM-300x166.png 300w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.34.42-PM-1024x565.png 1024w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.34.42-PM-768x424.png 768w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.34.42-PM-1536x848.png 1536w\" sizes=\"auto, (max-width: 1880px) 100vw, 1880px\" />\u003C/a>\n\t\t\t\u003C/div>\u003C/figure>\u003Cfigure class='gallery-item'>\n\t\t\t\u003Cdiv class='gallery-icon landscape'>\n\t\t\t\t\u003Ca href='https://lime-hamster-756747.hostingersite.com/?attachment_id=81'>\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"2348\" height=\"1154\" src=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM.png\" class=\"attachment-full size-full\" alt=\"Write a prompt to draft email to ChatGPT\" srcset=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM.png 2348w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM-300x147.png 300w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM-1024x503.png 1024w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM-768x377.png 768w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM-1536x755.png 1536w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Screenshot-2026-06-13-at-7.43.38-PM-2048x1007.png 2048w\" sizes=\"auto, (max-width: 2348px) 100vw, 2348px\" />\u003C/a>\n\t\t\t\u003C/div>\u003C/figure>\n\t\t\u003C/div>\n\n\u003Cp>\u003Cstrong>Step 3: Review and edit the draft\u003C/strong>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">ChatGPT will generate the result inside a writing block. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">For those unaware, writing blocks are a formatting feature OpenAI introduced last year. Instead of delivering a plain chat response, it allows you to review and edit parts of the result. You can also highlight a single line and ask for a targeted rewrite, as well as accept or reject individual changes one at a time. \u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>Step 4: Send the email\u003C/strong>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Once you’re satisfied with the draft, hit the send icon and ChatGPT will deliver the mail to the recipient. \u003C/span>\u003C/p>\n\u003Cp>\u003Cstrong>ALSO READ: \u003Ca href=\"https://www.neuralbriefly.com/blog/chatgpts-scheduled-tasks-makes-ai-do-the-work-while-you-sleep\">ChatGPT’s Scheduled Tasks makes AI do the work while you sleep\u003C/a>\u003C/strong>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">But there’s a caveat\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The biggest limitation to sending emails directly from ChatGPT is that it does not support file attachments. So, if your email requires PDFs, presentations, reports etc, you’ll need to stick to traditional methods. Another limitation is that the feature is only available for Plus, Pro, Business, and Enterprise plans. So ChatGPT Free and Go subscribers are out of luck—at least for now. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">However, OpenAI is still ahead of its rivals in introducing the ability to send mails directly from ChatGPT. \u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">Where this is heading\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Direct email sending is a relatively small feature but this version is helpful for text-only emails like scheduling meetings, sending quick updates, following up on a conversation, and responding to enquiries without leaving the ChatGPT window. Future versions could potentially handle attachments, manage inboxes, schedule follow-ups, support broader email clients, and do more tasks.  \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">For those who spend hours each week dealing with emails, this could turn ChatGPT from a writing assistant into a genuine productivity tool. It will not replace your email client entirely yet, but it will definitely save you a dozen unnecessary context switches every day.\u003C/span>\u003C/p>","2026-06-15T15:22:43","2026-06-18T23:39:02","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-15-2026-10_00_28-PM.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-15-2026-10_00_28-PM-768x432.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-15-2026-10_00_28-PM-300x169.png","June 15, 2026",3,{"id":59,"name":102,"avatarUrl":103,"description":104},[255,256],{"id":23,"name":24,"slug":25,"count":-1},{"id":65,"name":66,"slug":67,"count":-1},[258],{"id":32,"name":24,"slug":25},{"title":260,"description":261,"ogImage":248,"ogImageWidth":262,"ogImageHeight":263,"ogType":38,"breadcrumbs":264,"article":269},"ChatGPT can send emails directly: how to usse the new feature","ChatGPT can now send emails without leaving the chat. Here's how to set it up, what plans support it, and key limitation to know before you try.",1672,941,[265,266,267],{"name":41,"url":42},{"name":24,"url":44},{"name":243,"url":268},"/blog/how-to-send-emails-directly-from-chatgpt-without-opening-gmail-or-outlook",{"publishedTime":246,"modifiedTime":247,"author":102,"tags":270},[24],{"id":272,"slug":273,"title":274,"excerpt":275,"content":276,"date":277,"modifiedDate":278,"authorSlug":11,"featuredImage":279,"thumbnailImage":280,"smallImage":281,"formattedDate":282,"readingTime":16,"author":283,"categories":284,"tags":287,"seo":293},54,"chatgpt-memory-update-dreaming","ChatGPT’s new memory system solves the frustration of repeating context","No need to repeat the same details in every conversation.","\u003Cp>\u003Cspan style=\"font-weight: 400;\">After \u003Ca href=\"https://www.neuralbriefly.com/blog/chatgpts-table-of-contents-feature-announced-for-longer-conversations\">announcing an helpful table of contents feature last week\u003C/a>, ChatGPT is now getting a major memory update that can remember you properly. OpenAI is rolling out a new system called ‘Dreaming’ that better learns your preferences and keeps them fresh over time. It automatically updates memory in the background without you lifting a finger. So, if you have ever had to re-explain something like your travel plans, dietary preferences, certain work instructions every single session, that frustration now has a fix. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The latest update is currently live for Plus and Pro subscribers in the US, with Free and Go users set to receive it in the coming weeks and also expand to more countries.\u003C/span>\u003C/p>\n\u003Ch2>\u003Cspan style=\"font-weight: 400;\">Why ChatGPT’s new memory system matters\u003C/span>\u003C/h2>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">Until now, ChatGPT’s memory worked like a note system where someone took a few notes but forgot everything that wasn’t written down. You would explicitly have to say ‘remember this’. It often went stale and stayed wrong long after your situation changed. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The new ‘Dreaming’ system builds on the previous memory system launched in 2024. Instead of storing fixed notes, it automatically updates and organizes memory in the background, so the chatbot can understand what is still useful and what may no longer matter. The real benefit is convenience. A vegetarian user asking for meal ideas should get vegetarian options. Someone planning a trip should get suggestions that match their style. The system remembers things you’ve mentioned in past conversations.\u003C/span>\u003C/p>\n\u003Cdiv id='gallery-2' class='gallery galleryid-54 gallery-columns-2 gallery-size-full'>\u003Cfigure class='gallery-item'>\n\t\t\t\u003Cdiv class='gallery-icon landscape'>\n\t\t\t\t\u003Ca href='https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-scaled.webp'>\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"2560\" height=\"1707\" src=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-scaled.webp\" class=\"attachment-full size-full\" alt=\"Manage what ChatGPT should remember and what not\" srcset=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-scaled.webp 2560w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-300x200.webp 300w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-1024x683.webp 1024w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-768x512.webp 768w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-1536x1024.webp 1536w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Saved_memories.png-2048x1365.webp 2048w\" sizes=\"auto, (max-width: 2560px) 100vw, 2560px\" />\u003C/a>\n\t\t\t\u003C/div>\u003C/figure>\u003Cfigure class='gallery-item'>\n\t\t\t\u003Cdiv class='gallery-icon landscape'>\n\t\t\t\t\u003Ca href='https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-scaled.webp'>\u003Cimg loading=\"lazy\" decoding=\"async\" width=\"2560\" height=\"1707\" src=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-scaled.webp\" class=\"attachment-full size-full\" alt=\"Memory Settings page where you can turn it off.\" srcset=\"https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-scaled.webp 2560w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-300x200.webp 300w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-1024x683.webp 1024w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-768x512.webp 768w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-1536x1024.webp 1536w, https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/Memory_settings__1_.png-2048x1365.webp 2048w\" sizes=\"auto, (max-width: 2560px) 100vw, 2560px\" />\u003C/a>\n\t\t\t\u003C/div>\u003C/figure>\n\t\t\u003C/div>\n\n\u003Cp>The idea is simple: ChatGPT should not need the same context repeated every time. But this also raises an important question: what exactly does ChatGPT remember?\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">The updated memory system does not mean that ChatGPT remembers everything forever. There’s a memory summary page inside ChatGPT where you can see exactly what it knows about you – your interests, habits, preferences, etc. People can edit anything and delete details. You can access this by heading to Memory Settings and take a look at what’s there. \u003C/span>\u003C/p>\n\u003Cp>\u003Cspan style=\"font-weight: 400;\">With the updated memory system, OpenAI wants ChatGPT to become personal and userful over time. It makes memory controls more important for users who want to decide what ChatGPT should and should not remember.\u003C/span>\u003C/p>\n\u003Cp>Source: \u003Ca href=\"https://openai.com/index/chatgpt-memory-dreaming/\" target=\"_blank\" rel=\"noopener\">OpenAI\u003C/a>\u003C/p>","2026-06-06T15:02:28","2026-06-13T09:50:53","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-6-2026-at-08_30_03-PM.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-6-2026-at-08_30_03-PM-768x432.png","https://lime-hamster-756747.hostingersite.com/wp-content/uploads/2026/06/ChatGPT-Image-Jun-6-2026-at-08_30_03-PM-300x169.png","June 6, 2026",{"id":16,"name":18,"avatarUrl":19,"description":20},[285,286],{"id":23,"name":24,"slug":25,"count":-1},{"id":27,"name":28,"slug":29,"count":-1},[288,289],{"id":32,"name":24,"slug":25},{"id":290,"name":291,"slug":292},22,"Trending","trending",{"title":294,"description":295,"ogImage":279,"ogImageWidth":262,"ogImageHeight":263,"ogType":38,"breadcrumbs":296,"article":301},"ChatGPT can now remember your preferences better","ChatGPT's memory update now learns your preferences, stays current over time, and lets you control what it knows.",[297,298,299],{"name":41,"url":42},{"name":24,"url":44},{"name":274,"url":300},"/blog/chatgpt-memory-update-dreaming",{"publishedTime":277,"modifiedTime":278,"author":18,"tags":302},[24,291]]