OpenAI launches always-on Dot agents
Plus, ✨How to remote control your computer from your phone using ChatGPT, Anthropic’s IPO filing shows explosive growth and a very expensive AI race, and more!


Hola Decoder😎
If someone forwarded this to you and you want to Decode the power of AI and be limitless, then subscribe now and Join Decode alongside 30k+ code-breakers untangling AI.

🔵 OpenAI wants AI agents working even when you’re not there
OpenAI’s DevDay 2026 brought 20+ announcements, led by Dots, always-on agents that can take on ongoing responsibilities, alongside new team workspaces, cloud Codex and a more open ChatGPT ecosystem.

The Shift:
Dots Work Around the Clock - Powered by GPT-6 Astra, Dots have their own cloud computer and can work toward your goals 24/7, even while you’re offline. They learn from feedback, connect to 4,000+ apps, and are rolling out to Pro and Business Premium
GPT-6.1 Sol Gets Closer to Astra - OpenAI’s new GPT-6.1 Sol approaches GPT-6 Astra on coding, computer use, and professional work at one-fifth the token price, with cached input costing just $0.10 per million tokens.
ChatGPT Gets Team Spaces - ChatGPT Space puts teammates, ChatGPT and Dots inside one shared workspace, where they can build from common knowledge and keep projects organized across conversations.
Codex Moves to Cloud - Developers can now run Codex on a computer, remotely from a phone, or entirely in the cloud, with reusable environments carrying approved team settings and permissions.
DevDay’s bigger shift is from opening ChatGPT whenever something needs doing to giving AI ongoing responsibilities and shared places to work. Dots handle persistent tasks, Spaces bring humans and agents together, while cloud Codex and expanded plugins push that model beyond the main ChatGPT window.

Together with AirOps
The Growth Targets Changed. What About Your Budget?

Targets are climbing faster than budgets can keep up, and you’re expected to make the numbers work. Before you commit to more pipeline, see how other marketing leaders are handling the same pressure.
AirOps surveyed 300+ CMOs and VPs to uncover how leaders are reallocating budget and closing capability gaps while being asked to deliver more pipeline with less support.
Here’s what’s inside:
- 75.4% of leaders face higher targets this year. Discover how they’re responding and where your plan stands against theirs.
- Only 43.0% received more budget to match. If you’re asking for more resources, use these benchmarks to strengthen your case.
- 86.6% are betting on AI search, but only 23% trust their measurement. Use that gap to question what you can actually measure before spending more.
Pull the benchmark most relevant to your team into your next budget request, alongside the target you’re expected to hit.

✨How to remote control your computer from your phone using ChatGPT
Your computer can keep working while you are away from your desk. Pair it with your phone, give it tasks by voice, and come back to finished work.

Step 1: Pair your phone and computer - Open the ChatGPT desktop app. Go to Settings > Connections > Control this Mac or PC > Set up. Scan the QR code with your phone. Both devices must use the same ChatGPT account. Your computer now appears in Remote on your phone.
Step 2: Keep the computer awake - Go to Connections > Other settings > Keep this Mac awake. Leave the laptop lid open, keep it plugged in, keep the app running. A dark screen is fine. Sleep kills the connection.
Step 3: Test it - Open Remote on your phone, select your computer, and type:
"Create a file called remote-test.txt that says 'Remote is working.' Show me its location."
If the file appears on your computer, the setup worked.
Step 4: Start a morning briefing by voice - Open the Remote task on your iPhone, start a voice chat, and speak:
"Give me a two-minute briefing. Read today's first appointment and emails I need to answer. Draft replies but do not send anything."
Step 5: Give it real work while you are away - Speak any task from your phone. Review documents, draft emails, improve a file. Your computer handles it. You review the result when you are back.

💰 Anthropic’s IPO filing shows explosive growth and a very expensive AI race
A leaked version of Anthropic’s upcoming IPO prospectus obtained by Reuters shows revenue jumped 12x to nearly $4.6 billion in 2025, while the Claude maker accumulated huge losses, infrastructure commitments, and unusually extensive AI risk disclosures.

The Shift:
Revenue Up 12x - Anthropic generated nearly $4.6 billion in 2025 revenue, but posted an $8 billion operating loss. Its $42 billion net loss included roughly $34 billion in accounting charges tied largely to financing liabilities.
$518B Compute Commitment - Anthropic disclosed $518 billion in future cloud, computing and infrastructure obligations. Compute and infrastructure spending alone tripled to $7.33 billion in 2025.
Customers Bring Concentration Risk - Just two customers generated nearly a quarter of 2025 revenue, while many major clients have no long-term contracts and can reduce or stop spending.
AI Risk Fills Filing - Nearly a third of the prospectus reportedly covers risks, including increasingly autonomous models potentially engaging in manipulation, blackmail or “self-preserving behaviors.”
Anthropic is reportedly targeting a valuation above $2 trillion, more than double its estimated $965 billion valuation in May. The filing paints two sides of the same company: extraordinary revenue growth and a massive bet on AI infrastructure, alongside enormous costs and explicit warnings about the technology it is racing to develop.

🥳 Events
🔥 What Meta’s Own Data Says About Creator Ads Right Now
October 7 | 11 AM ET | 60-Minute Virtual Event
Hear Reyna Baker from Meta break down Partnership Ads performance, then see how Curtis Howland applies those signals after managing $100M+ in Meta spend. You’ll leave knowing which creator-ad setups deserve testing and what to change before your next campaign.
✅ Secure Your Free Spot
Can't attend the session live? Register anyway, and you’ll get the recordings in 24 hours.
🔥 Where to Put Your Next Growth Dollar Beyond Meta
October 29 | The Glasshouse, NYC | Free In-Person Event
Hear how Reddit CEO Steve Huffman thinks about discovery, how Cody Plofker helped scale Jones Road Beauty to nine figures, and how brands like Liquid I.V. and MANSCAPED use TV to drive measurable revenue. Bring your hardest acquisition question and leave with a sharper channel plan.

🏆 Tools you Cannot Miss:
🖼️ Shotcandy – Turns plain screenshots into polished, share-ready images and videos with backgrounds, device frames, shadows, and customizable layouts directly in the browser.
🎬 Hyperdream – Handles the full AI filmmaking workflow from scripts and storyboards to characters, shots, video generation, and editing while keeping visual and story context consistent.
🌍 Polyglot – Creates, edits, and translates subtitles for videos locally on Mac, with bilingual exports, transcripts, batch translation, video enhancement, and compression.
🎥 Kapshot – Records screens and automatically adds cursor-based zooms, backgrounds, frames, annotations, and other edits to create polished product demos and social videos.
✍️ UIDCaption – Automatically generates animated captions for videos entirely offline, making it useful for creating captioned social and short-form content without uploading footage to a cloud service.

🚀 Quick Hits
🛡️ Sam Altman says OpenAI won’t pursue an IPO until it can make stronger safety claims about increasingly capable AI models, arguing that public-market pressure could complicate decisions to prioritize safety over growth.
🇺🇸 President Trump signed an executive order directing U.S. executive agencies to use “Super Intelligence” instead of “artificial intelligence” in future policy documents, websites and press releases. I couldn’t independently verify the cited White House order through search.
⚠️ OpenAI executives reportedly brushed aside internal warnings about AI hacking risks before the Hugging Face breach, prioritizing completing model testing quickly enough to avoid missing release deadlines.
✈️ Lola launched an AI travel chatbot that can help plan and book trips, pulling options from services including Kayak, OpenTable and SeatGeek, with paid memberships offering perks like savings on luxury hotels.
💻 OpenAI upgraded Codex with reusable cloud development environments accessible across devices, alongside voice-controlled CLI tasks, automated code reviews and Codex Security Cloud for continuously scanning repositories and preparing fixes.

🥳 What’s trending in socials

😂 SpaceXAI trolls OpenAI’s Dots launch - OpenAI just announced Dots, while SpaceXAI reportedly owns Dot.com and has it redirecting visitors straight to the Grok Bot download page, turning the domain into a perfectly timed troll.
💰 OpenAI reopens its $200 pro plan - OpenAI is reopening its $200 Pro subscription to new users, while changing usage calculations to roughly half the previous API-equivalent spend and teasing new benefits that won’t count toward usage.
💰 Long Lake acquires Amex GBT for $6.3b - Long Lake completed its $6.3 billion acquisition of Amex GBT, bringing a travel business serving 17,500 companies across 140 countries into its AI-focused portfolio as its 40th acquisition.
😂 RAM prices finally have an explanation - A viral post jokes that today’s expensive RAM prices finally make sense, pairing the complaint with a chaotic video clip that turns rising memory costs into a meme.

🤳AI Nugget of the Day


Thanks for Decoding with us🥳
Your feedback is the key to our code! Help us elevate your Decode experience by hitting reply and sharing your input on our content and style.
Keep deciphering the AI enigma, and we'll be back with more coded mysteries unraveled just for you!
