In partnership with
Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?
Hi {{first_name|everyone}},
Figma wants AI to follow design rules, so today we’re looking at Figma Specs, a new way to give agents the guidelines before they start building.
Along with that, there’s a quick look at where design and code are blending together, a few handy tools, and a giant climbable spider made with a little help from AI.
Okay, let’s see what’s in store.
🤿 DEEP DIVE
Figma Wants AI to Stay on Brand
Giving AI agents a little more direction before they start designing.
AI can generate designs, write code, and build prototypes. But getting it to follow a company's actual design rules? That's a slightly different problem.
And apparently, that's one Figma wants to tackle.
At its first Config event in India, Figma introduced something called Figma Specs, a Markdown-based editor that basically gives teams one place to explain how their designs are supposed to work.
The idea is pretty simple. Instead of having AI agents generate something and hoping it follows the company's design language, give them the rules and context to work from in the first place.
Basically, a rulebook for humans and AI
Figma Specs lets teams write down their design guidelines in a document that sits either on its own or directly inside the Figma canvas.
But it's not just a document full of instructions.
Teams can reference actual designs, embed FigJam diagrams, include code layers, and connect their design and code libraries. So rather than simply telling an agent what a button should look like, they can give it access to the design elements and guidelines that define it.
And because everything is centralized, both human designers and AI agents can refer to the same information.
That's really the point here. Figma isn't just giving AI more things to generate. It's trying to give AI a clearer idea of what it's supposed to follow.
There's a reason they're doing this
Figma recently made its on-canvas AI agent available to more users. But there's an obvious complication with AI-generated work.
AI isn't deterministic. Give it instructions and you won't necessarily get the exact same result every time.
So Figma is building tools to make those results more predictable.
And there's a pretty revealing statistic from Figma's chief design officer, Loredana Crisan.
She says developers are producing 17 times more lines of code, but only 1.3x more software releases.
That's quite a gap.
Her explanation is that much of this extra code isn't necessarily meant to ship. It's being written to explore ideas and figure out what a product should actually be.
In other words, coding is increasingly part of the design process itself.
And Figma is connecting the two
Alongside Specs, Figma opened its code layers feature to public beta.
This lets users add code directly to designs on the canvas, creating more interactive prototypes. They can also ask an AI agent to turn a design into code or modify parts of that code to experiment with different ideas.
Figma is also integrating Payload, the headless content management system it acquired last year, into Figma Make. That brings content management into the process of generating dynamic websites.
Put those pieces together and you can see what Figma is working toward.
Design guidelines tell teams and agents what to follow. Code layers let them experiment with how something works. And Payload brings actual content into the websites they're building.
It's less about generating another design and more about connecting the different pieces involved in deciding what gets built.
The quick take
Figma Specs is a new Markdown-based editor for documenting design rules and giving AI agents shared context.
Teams can connect design libraries, code libraries, diagrams, and existing Figma elements to those guidelines.
Code layers are now in public beta, letting users add and edit code inside Figma designs with AI assistance.
Payload is being integrated into Figma Make to support dynamic websites and content management.
Figma's bigger focus is on making AI-generated design and code more consistent with what teams actually intend to build.
🤝 FROM OUR FRIENDS
A quick share we thought you might find useful
Yes, that Arnold Schwarzenegger.
He has a newsletter on beehiiv. So do Codie Sanchez, Jay Shetty, David Begnaud, Colin & Samir, and Joanna Stern.
They could publish just about anywhere. They chose beehiiv to build a direct connection with the people who want to hear from them most. And you can do the same, whether you’re world-famous or just at the start of your plan for world domination.
🤖 AI MADE THIS
Interesting and inspiring creations made with AI
A giant climbable spider designed between AI and human craftspeople
Theo Triantafyllidis used Gemini alongside clay models, sketches and traditional fabrication to design a giant soft spider that visitors can actually climb. Once onboard, they can enter a multiplayer virtual world built into the sculpture. The wonderfully strange project won this year's Frieze London Artist Award.
AI used
Google Gemini. Gemini helped generate sketches, character designs, backdrops and 3D renders. Photographs of real people were also converted into poses with Gemini before being turned into the game's 3D avatars.
⚡ QUICK HITS
The latest in AI, tech, and productivity worth knowing
Google Play is introducing a way to create personal Android apps using Google AI Studio and run them privately on your phone, with no developer account required. It's rolling out to adults in select markets, including the U.S., with the option to share your creations with friends and family coming in the next few months.
Cerebras CEO Andrew Feldman believes personal AI appliances aren't far away. He's picturing assistants that run partly on local hardware, learn your habits, and even have access to your credit cards to make purchases. He already uses a collection of AI agents to check his calendar and research people before meetings
👀 ON OUR RADAR
One more thing we thought you'd like
Your SOC 2, handled end to end with AI agents.
Enterprise buyers will not put your product near their customer data without a SOC 2 report. Sprinto gets you audit ready in 14 days, across three working sessions. AI agents do the collecting, you approve. Your auditor signs off.
⚒ TOOL SNAPSHOTS
Futuristic tools within AI, no-code, and productivity
Get an AI agent to handle tasks across websites.
Why it’s useful: Handy for automating repetitive browser work, like finding leads, reaching out to prospects, or applying for jobs, without doing every step yourself.
💻 Inferock
Run different AI models and keep track of what they cost.
Why it’s useful: Useful for developers testing multiple models who want to monitor spending, spot failed API calls, and recover credits on eligible failures.
Take some of the daily work out of social media marketing.
Why it’s useful: Gives small marketing teams a way to research content ideas, create videos, plan posts, and spot relevant conversations without juggling separate workflows.
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?
