In partnership with
Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?
Hi {{first_name|everyone}},
OpenAI already makes more than half a billion images a day, so the next step was getting them to animate, and Images 2.6 can turn even a rough sketch into a moving character. Along the way, we’ve got a quick share from our friends at Apper, a couple of thoughts on how AI is changing work, some handy tools, a bit of internet history, and one pretty cool AI creation. Let’s start with the moving pictures.
🤿 DEEP DIVE
From Still Images to Moving Ones
OpenAI adds animation, faster generation, and finer control.
Apparently, generating half a billion AI images a day wasn't enough. Now OpenAI wants to make them move.
Okay, so OpenAI has introduced Images 2.6 across ChatGPT, Codex, and the API. And while there's the usual stuff about faster generation and better image quality, the addition that caught my attention is GIFs.
You can take a simple sketch and turn it into an animated character. Or make a brand mascot move, create a looping product shot, or put together animated storyboards to explore how a character moves before making a video.
And you don't need to start with some beautifully finished illustration either. Even a rough sketch can become an expressive animated character.
That's a pretty fun addition.
But there's more going on here
The other big improvement is speed. Images 2.6 generates images 50% faster than before, which is especially useful when you're making something that needs several rounds of edits.
For developers, OpenAI is offering two models through the API.
Flare is the faster option. At low quality settings, it can generate an image in as little as six seconds. It also improves layouts for things like posters and product visuals.
Sunburst is more focused on detail and consistency. It's designed to keep characters, products, and other visual elements looking the same as you make changes, particularly at medium and high quality settings.
And I think that consistency part is easy to overlook.
If you're working on a series of visuals featuring the same character or product, you don't really want it changing appearance every time you ask for an adjustment. The idea is to keep what you've already created while refining the bits that need work.
A few smaller details worth noticing
Image markup, which was already available in ChatGPT, is also coming to the API. That lets developers point to specific areas of an image and indicate exactly what needs changing.
There's also SVG export for designs that need to work across websites, apps, or print.
And developers get a new complexity setting, which controls how much planning the model puts into a request. That's separate from image quality, which controls visual detail. So a quick edit doesn't necessarily need the same level of planning as a more complicated scene.
One number that puts all this into perspective is that people are already creating 4 billion images every week across ChatGPT, Codex, and the API.
That's more than half a billion a day.
So while GIFs are probably the most playful part of this update, the bigger idea is making image creation less about generating something once and more about being able to keep working on it, adjusting it, and even animating it.
And all of this is available globally now, with Images 2.6 rolling out across every ChatGPT and Codex subscription tier.
The quick take
Turn sketches, characters, mascots, and product visuals into animated loops.
The Flare API model can produce low-quality-setting images in as little as six seconds.
Sunburst focuses on visual quality and keeping characters and products consistent across changes.
Image markup, SVG exports, and a new complexity setting offer more ways to refine results.
Images 2.6 is available in ChatGPT and Codex across all subscription tiers, with Flare and Sunburst in the API.
🤝 FROM OUR FRIENDS
A quick share we thought you might find useful
Build Real Apps Inside Claude or ChatGPT
Meet Apper, an AI app builder that provides unlimited AI messages to build with ChatGPT or Claude.
No need to buy AI credits or worry about hitting limits anymore!
Apper provides a ready-to-use backend with a database, user logins, and hosting for your vibe-coded apps.
It also adds a second layer of security to your AI-generated code, with features like:
Private/Public access: Your data is private by default. You choose what goes public.
User profiles and permissions: Each user can only see or edit the information they’re allowed to access.
Development mode for your apps: Apper lets you separate development and production apps, including separate databases. So you can safely make changes without affecting your live app.
With Apper, you can even publish your app in 1-click or add your own custom domain.
Free to start. No card needed.
🔥 WHAT’S THE BUZZ
What people are building, sharing, and talking about
AI researcher François Chollet shared an interesting observation about AI-assisted software development. He argues that while coding agents increasingly handle the actual programming, software engineers are still working hard, and the financial results of companies using these tools aren't necessarily improving at the same pace.
This applies beyond software development. AI might help someone produce ten reports instead of two, but that doesn't necessarily mean the company makes better decisions.
The useful question is whether AI is improving the outcome or simply increasing the amount of work being produced.
Felix Rieseberg says he has stopped using traditional coding interviews to hire engineers for his team at Anthropic. His reasoning? AI can already solve many of the coding challenges companies use to evaluate candidates.
The idea extends beyond software engineering. As AI makes execution easier, being able to identify problems, propose useful solutions, and demonstrate results could become increasingly valuable at work.
⚒ TOOL SNAPSHOTS
Futuristic tools within AI, no-code, and productivity
🤖 Tobira
Turn AI chat traffic into leads by making your site an agent.
Why it’s useful
It answers questions, negotiates, books calls, and takes orders inside ChatGPT, Claude, Grok, and other AI chats. Leads can land in your Telegram or email, which is handy if you want follow ups without living in your inbox.🧠 Sage2
Build apps that pull structured data from conversations.
Why it’s useful
It helps developers extract dates, amounts, and durations from natural language and decide what to do as a conversation changes. Useful for apps that need to track shifting user intent and pick the next step without hard coded rules.🛡️ GPT-6 Cyber
Helps security teams research and test cybersecurity risks.
Why it’s useful
Use it when you need deeper security research with tighter safety guardrails. It’s gated for vetted Daybreak Red customers, so it fits teams that can handle identity checks, MFA, and an application.
📜 FROM THE ARCHIVES
A look back at a moment that shaped tech, AI, or the internet
May 3, 1978: The first spam email was sent
Gary Thuerk, a marketing executive at Digital Equipment Corporation, sent an unsolicited promotional email to hundreds of ARPANET users. The reaction was immediately negative, but the message also showed how email could be used as a mass-marketing channel long before inboxes, spam filters, and email etiquette existed.
👀 ON OUR RADAR
One more thing we thought you'd like
Vanity Fair’s New Show, Where No Subject is Off Limits
Join Mark Guiducci for a series of in-depth and far-reaching conversations with the most interesting characters from the worlds of Hollywood, Politics, Style, Silicon Valley, and beyond. It's all Fair Game.
🎁REFERRAL REWARDS
If TIW has been useful, invite one friend or coworker who’d actually use AI/automation at work.
1 referral → get the TIW No-Code Workflows PDF (25 plug-and-play workflows)
10 referrals → get a full course free (your choice)
Your referral link: {{rp_refer_url}}
Tip: send it to a friend who’s AI-curious, a coworker who wants to move faster, or anyone trying to automate repetitive tasks.
🤖 AI MADE THIS
Interesting and inspiring creations made with AI
Rehan Sheikh and the Patch team used Claude Opus 5.5 to create hundreds of animated pixel art effects, from fireballs to lightning strikes. Instead of generating images, Claude wrote and refined the animation code. The collection is free to explore, and its GitHub repository has already earned thousands of stars.
AI used:
Opus 5.5
ℹ️ ABOUT US
The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.
We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.
Our mission is to empower individuals, boost their productivity, and future-proof their careers.
We read all your comments - please provide your feedback!
Did you like today's email?
What more do you want to see in this newsletter?
