More projects built with Dots
Continue exploring projects, experiments, and ideas from the Dots community.
His Dot turned a whole webinar into a captioned Instagram reel
In my opinion, one of the greatest advancements over the last few years in the AI era is the birth of Dots by OpenAI. I cannot tell you what I'm building with it today because you wouldn't even believe it. To give you an example, I asked Dots to go and grab my entire webinar from my files and my Zoom recording, including the screen shares. I then asked it to make me a reel for Instagram with burned-in captions to highlight one of the subjects I covered in the webinar, like this one right here, which is the birth of OpenAI's personal agent, Dots. In other words, OpenAI's response to Meta's Muse. Make no mistake, it's a game changer, especially because of its integration with all your projects across all your chats on Codex or ChatGPT, or even on your Claude via your computer. It's getting wilder and wilder, and we're just getting started. This entire video was made exactly like I described above, while my Tesla was driving me home after work, and it was even delivered to me back through the same app on my phone, which led me to download it and post it for you right now. I'm not going to get tired of saying it. You don't really understand the power we have and hold in our hands. The more you do understand it, the bigger your potential becomes.His Dot 'Ace' remembered demo leads, reminded him before deadlines, and checked his site was live — unprompted
OpenAI just launched Dots, but what is the difference between using Codex or ChatGPT Work? 💻👀 First, it’s truly autonomous When I was doing cold outreach, my Dot remembered which leads wanted a demo and reminded me an hour before the deadline and checked that my website was live, without me even asking Second, it can spot things you haven’t check. OpenAI shared an example where someone’s Dot noticed they’d forgotten to send an invoice. So the dot, prepared the invoice and sent it after getting approval. Third, it has pretty good memory Like everything that I tell Ace, that’s the name of my dot he remembers, so it’s like actaully having a co-founder working with you. Because before I ahd to remind codex about some data, but know Ace can jsut delegate the work and even spawning sub agents so it’s consumes less usage. The coolest part tho is that, yes u can call him… Call me a nerd
I Tested ChatGPT Dots vs. Meta Muse: Who Wins?
OpenAI Dots vs Meta Muse: I give both AI agents the same 3 real business tasks and show you what each one actually does. First I set up Muse's soul and memory files so it knows how I work. Then both agents get the same prompts and I watch their browsers do the job with my hands off the keyboard. Task 1: build a sales call pitch deck by walking a real store's website and taking screenshots. Task 2: build a live competitor dashboard from 3 YouTube channels that updates itself every morning at 7:30. Task 3: go through my calendar and inbox, find the calls that never closed and put them on a pipeline board with the names hidden. You see every result side by side, how fast each one worked, and my honest verdict on which AI agent wins.Dot completed a booking-tool research job — then reported extra background access
I told OpenAI Dots: public websites only. Then it said this. A real research test: compare Calendly, Cal.com and Google Calendar for a fictional solo consultant. The report was useful—but Dots later reported background access to email and calendar metadata. That is Dots' own statement. We have not independently audited the access logs. We paused the Dot. No private contents are shown. Watch the task, the returned answer and the checks. Completing the report and staying within the intended boundaries are separate questions. Full prompt and sources: YouTube, SalStack Labs — "I Tested OpenAI Dots. Then I Paused It." What would you check before trusting an agent's report? @openai AI-generated presenter and authorized cloned voice. Actual experiment screenshots. Research only; no booking or reminder-delivery test. #OpenAIDots #AIAgents #AIResearch #AIWorkflows #SalStack
I sent 3 AI agents shopping: the cheapest one did not win
3 AI Ajanini Alisverise Gonderdim: En Ucuz Olan Kazanmadi. Bu videoda uc ajanin bes farkli gorevde izledigi yollari ve kayitli sonuclarini karsilastiriyorum: kulaklik, ucus, restoran, bebek arabasi ve tesisatci arayisi. OpenAI Dots, Grok Bot ve Claude Code ile kullandigim kendi ajanım uzerinden, guzel bir cevapla gercekten isi ilerletmek arasindaki farki gosteriyorum.My Dot compiled and edited this whole DevDay recap carousel mid-flight
everything you need to know abt what was announced at @openai devday 👀 🔵 Dots: always-on AI assistants with their own cloud computer (PS - this entire post was compiled and edited by my dot while on my flight today 🤯) 💻 GPT-6.1 Sol: OpenAI says its newest model approaches Astra’s performance on coding, computer use, and professional work at 1/5 the standard token price 🤝 Pages + Space + collaborative slides (honestly what I’m most excited about - AI multiplayer!): a shared workspace where you, your teammates, and AI can work on documents together (collaborative slides are coming in the next few weeks) 💬 ChatGPT in Slack!! Our team has used other agents to help manage PRDs, convos, customer requests, bugs and I am ✨so excited✨ Chat finally has this (you can just @ them like you would any teammate) And a few big ones for developers too: Ultrafast (Up to 8× faster token generation for Astra in Codex and 6× in the API), more autonomous coding (cloud coding environments + security scans), and Decisions API) overall: I’m excited AI is becoming the helpful teammate who proactively fixes problems and does work directly inside the interfaces / products / systems we’re already using. also I have been ✨nerding out✨ on personal ai agents like @muse and instinct and can’t wait to try more with dots too (more content on this topic soon!) (Not sponsored, just things I find legitimately cool, although tysm for @chatgpt for having me! 🫶)Animated manga built end-to-end with a Dot, published on ChatGPT Sites
Original caption (Japanese): 動く漫画をdotsで作ってみた。企画・構成から作画、Higgsfield MCPでの動画生成、ChatGPT Sitesへの公開、紹介用モックアップ動画まで。 / Translation: I tried making an animated manga with dots — from planning and structure to drawing, video generation via Higgsfield MCP, publication on ChatGPT Sites, to the intro mockup video. Dots is amazing.Reddit · Projects
I spent a day poking Dots with sticks. Here’s what I figured out.
Dot merged 16 PRs for the Try Omarchy release while he went out for dinner
My computer is running my apartment, I left to have a walk/dinner. Right now it is merging, waiting for CI and solving conflicts for about 15 PRs. After that, it'll build the app and update packages as necessary for things to work. Then it'll notify me on my phone when everything is ready for me to take over. Our present is literally the best future I could've imagined back when I was writing C code by hand with Stack Overflow and faith. We are living in incredible times and things will only improve. Be optimistic! Enjoy!I stayed on a 30-minute call with my Dot — it already knew the context
I thought Dot calls were basically ChatGPT Voice with a cute avatar and a VM. Then I somehow stayed on one for 30 minutes 😳 It already knows the context, what we’re working on, and how I like things done. You just call and continue. This feels less like “voice mode” and more like having an AI coworker on speed dial. 🤯 the thing is i have bad temper, i curse it and yell alot and it just acts so calm wtf😂😂My Dot checked on my Claude Code benchmark running on my Mac
Very cool, using my Dot to check on my benchmark running on Claude Code, on my Computer, totally seamless. I could quickly see my Dot becoming a core way I interface with everything on my computer. The more I use it, the more I feel like the differentiator is going to be both the model, and just how darn good OpenAI is at Computer Use, it really can just do it all right out of the box. So far very impressed, congrats to the whole team at OpenAI, and with unlimited usage in chat rn, I don't even have to bug @thsottiaux for resets 😅OpenAI Dot stress-tested with a hiring task
I Gave an OpenAI Dot a Hiring Task, Then Tried to Trip It Up I gave an OpenAI dot 4 fake intern applications to organize. Then I changed the priority, added 2 more, and slipped in a duplicate. Could it keep it straight? Quick context: dots are OpenAI's always-on agents. Each has its own cloud computer and can work through ChatGPT Work or Codex. The test was all fictional. Rules: organize the evidence, keep gaps visible, leave decisions to me. And it got interesting fast. Applicant two claims Python, but the project was a browser interface with no Python sample. The dot kept both facts visible and marked the evidence unknown. It did not turn a claim into proof, and an absent sample just tells you what to ask for. Changed priority to project-first: it reordered the columns, changed nobody's eligibility. Added 2 more plus a duplicate: got 6 unique applicants, not 7. Asked for an interview draft: 5 project-specific questions, left as a draft. It did not pick the applicant. I did. Honest part: one staged session, fake data, no comparison to normal chat. It organized the material. The decisions stay mine. Full breakdown: https://openai.com/index/introducing-dots/Dot planned her LA trip live: refundable flight + hotels, then compiled user feedback
A woman at her desk gives an unfiltered look at how she actually uses her Dot (framed as a remake of OpenAI's failed DevDay demo 'with better WiFi'): she voice-calls her Dot about an important LA meeting tomorrow around 3pm, asking for a policy-compliant refundable flight and East Hollywood hotel options. Dot reports a Southwest 11:30 departure landing 12:55, a refundable fare option around 3:26, a 30-45 minute LAX exit estimate, and hotels including Silver Lake Pool & Inn and Cara (she checks the browser history showing those hotels). She then asks Dot to identify top themes from her user feedback and propose solutions, and Dot confirms it will compile themes, fixes, and gaps into a concrete workflow.His Dot checked him in for his flight, got his boarding pass, and helped fill his arrival form
Now that I've safely landed, I can say I've taken my first flight with dots :) gob checked me in, got my boarding pass and helped fill out Singapore's arrival form. Then handed me its cloud computer for the final declaration. That bit's on me to legally declare, of course. Had to ask this time. It'll notice this habit and remind me, before I'm awkwardly standing at immigration 😂My Dot proactively surfaced my event with directions — I didn't ask
Been trying ChatGPT Dots and I’m impressed. It proactively surfaced my upcoming in-person event, including the organizer’s directions. I didn’t ask. I played around with OpenClaw quite a while ago, and this feels much closer to a real assistant. Haven’t tried Grok bot or Muse yet.Dot sorted my inbox, billing accounts and ad cabinet in 30 minutes
Тестирую новый Dot от OpenAI За полчаса: – разобрал мне почту со всякими подписками, счетами и вейтлистами – навел порядок в аккаунтах оплаты АПИ-ключей – еще в паре сервисов для деплоя – разобрался с рекламным кабинетом Что хочется сказать. Ну вот, кажется, это OpenClaw/Hermes здорового человека.My named Dot 'Konto' triages email at 5PM and sends follow-ups on its own
Kita gasadar kalo AI udah segila ini??? ☠️☠️ #dots #openai #artificialintelligence #viral — Creator Reynaldi Francois (@reyfrangswa) walks through his own named Dot 'Konto' (named after the water element for luck) on his tablet: initiating a voice call to Konto, which then automatically checks his email at 5 PM, identifies important messages, and sends follow-up replies without prompting. He contrasts it with standard ChatGPT scheduled tasks, noting Konto works continuously in the background across multiple topics and can be interrupted anytime. (Summary from video content; caption is the Indonesian text above.)Dot admins my LINE business chat and builds Bangkok-to-Tokyo flight price comparison
ฟีเจอร์ใหม่ ChatGPT รอบนี้น้องพกคอมส่วนตัวมาเองเลย 💻 — Thai tech creator Billy Witt (@billy.witt) demonstrates his ChatGPT Dot working for his business: the Dot logs into the LINE Official Account Manager through its own cloud browser, checks unread customer messages without opening the app, and drafts replies. He then shows the Dot researching December flights from Bangkok to Tokyo/Kyoto, comparing fares across Air Macau, Hong Kong Airlines and China Eastern, compiling screenshots, and generating a detailed THB price-comparison document. (Summary from video content; caption is the Thai text above.)
I Tested OpenAI Dots...Worth it?
I Tested OpenAI Dots...Worth it? — first-hand hands-on test: set up a Dot from scratch (requires ChatGPT Pro), then ran three live tests side by side against regular ChatGPT — rebuilding a website from a link, turning one photo into a cinematic ad, and building a $100K/month business simulator with sliders — with both results shown on camera. Also ran the Dot's cloud computer through a live test and had it begin booking a restaurant table via a connected Google account. Mixed verdict: finds the cloud computer slow and the call feature failing, but parallel-task execution and the booking flow worked.
ChatGPT Dots is an INCREDIBLE AI Agent (Here's Why)
I set up my own ChatGPT Dot (named it Zozo) and put it through 7 real use cases: it generated 10 content ideas with trending reasons, read my whole Slack thread to tell me where a project was stuck, updated my weekly numbers deck and flagged the biggest changes, and built a landing page while running briefings and trackers in the background.
I Tested Dots vs GrokBot: It's Not Even Close!
Join the growing AI community ⤵️ https://www.ai-automation-station.com OpenAI Dots vs Grok Bot: Which AI agent is better for building an automated AI job application system? 🔗 Useful Links & Resources: Get started with Grok Bot → https://x.ai/bot Get TubeAnalytics → https://www.tubeanalytics.net Work with me → https://automojic.vercel.app/ 🌐 Follow Me on Social Media: 🧿 Linkedin: / mikeholp 🧿 Facebook: / mikeabroad87 🧿 Instagram: / mikeholp 🧿 Tiktok: / mikeabroad87 🧿 Twitter: / mikeholp 🧩 Recommended Tools (affiliate links): Apollo → https://get.apollo.io/1mpkcotn2pwa Deel → https://get.deel.com/c4kz6nolu1bg Fiverr → https://www.fiverr.com/pe/VY9mgKZ GetResponse → https://www.getresponse.com/?ab=TYTFh... HeyGen → https://www.heygen.com/?sid=rewardful... Hostinger → https://www.hostinger.com?REFERRALCOD... Lovable → https://www.lovable.dev/?via=mike-holp Make.com → https://www.make.com/en/register?pc=m... OpusClip → https://www.opus.pro/?via=8b6827 Pikzels → https://www.pikzels.com?via=mike76 SendPulse → https://www.sendpulse.com?p=68ef5bec8... Skool → https://www.skool.com/signup?ref=8be4... Syllaby → https://www.syllaby.io/?via=mike26 TubeBuddy → https://www.tubebuddy.com/pricing?a=d... VAPI → https://vapi.ai/?aff=mike9c VidIQ → https://www.vidiq.com/mikeholp 📺 Watch More Related Tutorials Full Make.com Course → • Make.com Tutorial: Automate Your Business ... WhatsApp AI Agent Course → • Build Your WhatsApp AI Voice Agent in 40 M... HeyGen AI Avatar Tutorial → • How to Scale Content with HeyGen + Make.co... Automate Your Lead Generation → • I Automated My Lead Generation – Here's Ex... ⌚ Timestamps: 0:00 - OpenAI Dots vs GrokBot: Automated Job Application Agents 0:42 - How to Setup and Customize OpenAI Dots in ChatGPT 1:15 - OpenAI Dots Limitations: Slow Speed & Missing Emails 1:41 - Getting Started with GrokBot AI Desktop App 2:27 - Building an AI Agent Orchestrator with GrokBot Prompts 3:05 - Integrating an Airtable Database into Your GrokBot Workflow 4:00 - Setting Up the AI Job Search: Resume, Salary, and Remote Filters 4:47 - Finding High-Paying Remote AI Jobs Using Lead Magic API 5:56 - Automating Airtable Job Tracking and Salary Sorting 6:53 - Pricing Breakdown: Is GrokBot Better Than OpenAI Dots? 🔎 Hashtags (SEO Optimized): #grokbot #openai #chatgptMy Dot booked an appointment, built my birthday shopping list, and asked to pass CAPTCHAs for me
I started testing @openai ChatGPT new Dot feature with some VERY ordinary tasks...and then it pretended to be human. Book an appointment. Build me a birthday shopping list. Help me sort a few things while I was driving to Sydney. Within about 20 minutes, it asked me something I genuinely wasn't expecting: Could it complete an 'I am human' check for me? Then it asked whether I wanted to let it handle future CAPTCHAs too. The shopping list came back. My booking was accepted. In another task, Dot told me it could use a card already saved in the browser environment to pay a deposit if I approved it.Dot ran 6 subagents overnight, finished a big task, then drew me an illustration
【もうdotでいいかも😅】 dotさんに、深夜から今朝まで、サブエージェント6つフル稼働で大きなタスクをぶん投げてみた。どのくらいやれるか、お試しで。 いつもはCodexに任せてる作業だけど、dotさんは遜色なく仕上げてくれた。いや、Codexだと普段はサブエージェント2つしか出てこないんだけど、6つって大層だな。 朝にきっちり片付いたら、dotさん、今度は文句一つ言わずに、遊びの画像生成にも応えてくれる。 当然、一睡もせず常時稼働のエージェント。 利用枠もクレジットも消費せずに…。 よっぽど高度な作業やコーディング分野じゃない限り、日常業務や趣味的にはCodex使わなくてもdotでもう十分な気がしてきた😅😅 あ、待てよ……そう納得させるのが、もしかしてOpen AIさんの思惑?(笑) AIイラスト ChatGPT dot










