It's Wednesday, September 9th. Meta shipped an agent that pays with a one-time card and asks before it sends. OpenAI's Images 2.5 reaches every tier, free included. Below, the two prompts I ran side by side today, and which one produced the picture you would use.
In Today’s Top Tools, we spotlight some of the most innovative, creative AI apps that we recommend adding to your stack.
TOOL SPOTLIGHT
1️⃣ Meta's Muse Runs Your Errands From Inside A Locked-Down Cloud Computer
Image from Meta
Muse is Meta's personal AI agent for people who want tasks finished rather than chatted about. It lives on its own cloud machine with a second agent, Sentinel, guarding every outbound action. The containment is the product pitch, though Meta has published no prices for its paid plans.
Meta started rolling out Muse on September 8, a personal agent that books travel, fills forms and pays at checkout while a separate gatekeeper decides what reaches the internet. It runs on Muse Spark, Meta's model built for agentic work.
Message Muse in WhatsApp or its own app to lower a bill, sell a car, or plan travel.
Pay through Link built by Stripe, which issues a one-time-use card so your real card details stay hidden.
Review a full audit trail of what Muse did, and tell it to forget any detail it learned.
Free for most of what people need with subscription plans on top, US only for now, on iOS, Android and muse.ai.
Muse is a consumer launch, but the people it will surprise are the ones on the other side of the checkout. It opens a browser, fills out forms, negotiates on someone's behalf, and pays with a one-time-use card from Link. If you run a cart or a support inbox, an agent will be working it soon, and it reads your page far more literally than a person does. The useful hour this month goes to walking your own funnel the way an agent would see it. Worth stealing too, Meta put a second agent in front of the first one. That pattern travels.
TOOL SPOTLIGHT
2️⃣ ChatGPT Images 2.5 Keeps Your Edits From Falling Apart Halfway Through
Image from OpenAI
OpenAI's image upgrade is aimed at anyone who edits a picture more than once. Earlier changes hold their shape across a long conversation, and subjects from a reference photo stay recognizable. It ships to every tier including free, so the test is whether the fifth edit still looks right.
OpenAI started rolling out ChatGPT Images 2.5 on September 8, with steadier multi-turn editing, a Sketch tool that turns a doodle into a reference, and two API models led by GPT-Image-2.5 Flare.
Draw a rough shape with Sketch, type @Sketch in ChatGPT, and let it guide the finished image.
Start from a Poster or Merch template instead of a blank canvas, then share the prompt alongside your result.
Build with Flare for everyday volume at 50% lower latency, or Sunburst for production campaign edits.
Live today on every ChatGPT, ChatGPT Work and Codex tier including free, with API image output at $30.00 per million tokens.
This one reaches the free tier too, which means the constraint stops being access and starts being brand consistency. Everyone on your team can now build a product shot up across a long conversation instead of regenerating from scratch, and the edits mostly hold. Put it where the stakes are low first, mockups and social. Save the campaign work for Sunburst with a real reference photo in the loop. The prompt-sharing feature is the quiet one to use internally, because a shared prompt is the closest thing you have to a style guide people will actually follow.
🎓AIC AUTOMATIONS
🛍️ Make Your Own Product Shots And Ads
Images 2.5 shipped yesterday to every ChatGPT tier, free included. If your business needs a product photo or a poster, make one today.
Every business needs two pictures, a clean shot of what you sell and an ad you can post. Both came from a public prompt library, ninety seconds each.
Case 15, Spicy Chili Chutney Product Shot, by @iamaiistudio.
Four Moves
Pick the case closest to what you sell.
Swap in your product and colors.
Put your label text in quotes. An unnamed label comes back blank or invented.
Change one thing per turn. Images 2.5 holds earlier edits more reliably.
Ultra-realistic premium food advertisement poster for a healthy breakfast granola brand, centered matte pouch packaging labeled “Matcha Oat Granola”, green monochrome aesthetic, flat lay composition, soft studio lighting, vibrant matcha green background, surrounded by kiwi slices, almonds, oats, chia seeds, matcha powder bowl, granola bowls, scattered ingredients, clean modern typography headline “SUPERFOOD MORNING BOWL”, handwritten annotation arrows with wellness benefits, luxury organic branding, natural shadows, high-end commercial food photography, minimal yet detailed layout, symmetrical composition, sharp focus, Instagram ad style, 8k detail, healthy lifestyle marketing design.
Swap these. Granola brand, matte pouch, “Matcha Oat Granola”, matcha green, “SUPERFOOD MORNING BOWL”.
Case 23, Matcha Granola Ad Poster, by @Sairah_0. The pouch reads Verdura Organic Foods, a brand the model invented, so name yours.
Ideas Worth Stealing
462 examples, free to reuse, all credited.
E-commerce. Hero shots for a listing page.
Ad Creative. Layouts with room for a headline.
Portrait & Photography. Lighting words for a founder headshot.
Pro Tip. I made both with Codex CLI on my ChatGPT account. Run codex exec -C <folder> -- "<prompt>, save as picture.png" and the PNG appears. No API key, no per-image bill. Codex is included across ChatGPT plans, Free included, limits by plan, and it needs a terminal.
One way to put AI to work, every week. From Joy Dong
In this section, we feature a few standout opportunities at leading AI companies, non-profits, policy groups, and other organizations.
🔬 Work at a Lab/AI Safety
Roles at frontier labs, safety organizations, and research institutions
RESEARCH SCIENTIST/ENGINEER (EVALUATIONS) - Apollo Research | London & San Francisco | 100k-200k GBP (~150k-270k USD)
RESEARCH ENGINEER/RESEARCH SCIENTIST, RL/REASONING - OpenAI | San Francisco, CA | $295,000-$445,000 + Equity
RESEARCH ENGINEER - FAR.AI | Berkeley, CA / Remote (International/US) | $100,000-$190,000
MEMBER OF TECHNICAL STAFF - MID-TRAINING - xAI | Palo Alto, CA | $180,000-$440,000
🚀 Work in Industry
Roles at funded AI startups and private companies
STAFF PRODUCT MANAGER, ML FOUNDATIONS AND GENAI - Stripe | New York, NY / Seattle, WA / South San Francisco, CA / Remote US | $214,300-$321,500
SENIOR / STAFF PRODUCT ENGINEER, AI - Linear | Remote (North America) | Competitive Salary + Equity
SR. PRODUCT MANAGER, DATABRICKS AI - Databricks | Seattle, WA | $148,800-$204,525
SOFTWARE ENGINEER, AGENT EVALUATION AND QUALITY - Cursor (Anysphere) | San Francisco, CA / New York, NY | Competitive Salary + Equity
MEMBER OF TECHNICAL STAFF (SOFTWARE ENGINEER, APPLIED AI) - Perplexity | San Francisco, CA / New York, NY / Palo Alto, CA | $220,000-$405,000 + Equity
🏛️ Work in AI Policy / Governance
Roles in government, think tanks, NGOs, and policy organizations
AI POLICY RESEARCH RESIDENT, ASSOCIATE / SPECIALIST - RAND Corporation | Washington, DC | $110,900-$183,300
DC RESEARCH MANAGER (TALENT DEVELOPMENT TEAM) - GovAI | Washington, DC | $105,000-$150,000
POLICY ADVISOR / SENIOR POLICY ADVISOR - Encode AI | Washington, DC | $110,000-$180,000
✨AIC OPPORTUNITIES
Current ways to build with The AI Collective.
DIRECTOR OF SF BAY AREA PROGRAMS - The AI Collective | San Francisco Bay Area | Paid, full-time, $60,000-$80,000/year
EDUCATION OPERATIONS LEAD - The AI Collective | Remote | Volunteer, 5-10 hrs/wk
EXECUTIVE ASSISTANT - The AI Collective | Remote, LatAm preferred (PST hours) | Paid, $1,500/month
PARTNERSHIPS LEAD (GRANTS) - The AI Collective | Remote | Volunteer, part-time, ~5-7 hrs/wk
🫵 Want your message in front of 200,000 AI builders?
Our partners and sponsors get exclusive placements across the newsletter and access to AIC's in-person network — demo nights, dinners, hackathons, and forums across 180+ chapters.
For all inquiries, send us a note at [email protected].
The AI Collective is built by volunteers across 180+ chapters in 40 countries.
Thank you to the thousands of volunteers around the world who make this work possible. We truly could not do this without you.
🧑💻 About the Editors
About Noah Frank
Noah is a researcher, innovation strategist, and ex-founder thinking and writing about the future of AI and the workforce. His work and body of research explores the economics of emerging technology and organizational strategy. Outside of AIC, Noah heads research for Centaurian AI.
About Lindsay Gross
Lindsay is an AI engineer, researcher, and writer focused on how AI systems behave in practice and what it takes to make them safe. Her work sits at the intersection of AI safety, governance, and product design, and at AIC she writes about the questions that matter most as these systems scale.