Skip to main content

Natural language prompts

2022, ChatGPT

In 2017, eight scientists at Google Brain published “Attention Is All You Need,” a research paper that lay the foundation for today’s LLMs by introducing deep learning architecture known as the transformer. Transformers allowed AI models to read text out of order, decide what’s important, and draw meaning based on relationships instead of position. The breakthrough enabled OpenAI to launch ChatGPT in 2022, now joined by Google DeepMind’s Gemini, Anthropic’s Claude, Meta’s Llama 2, and many others. By allowing users to interact with machines through casual conversation

Cooking with constraints: A designer’s framework for better AI prompts

Design and cooking share a truth: Preparation determines the outcome. Structured prompts turn AI from guesswork into a reliable design partner.

, natural language prompts made it possible for virtually anyone to generate anything—regardless of their existing knowledge or technical skill.

This has opened up a new way to relate to software—not just as a tool, but a companion

Cover illustration of a comic showing a person walking a large fluffy dog through a colorful, surreal cityscape with the title “Hard Problems Are Still Hard.”Cover illustration of a comic showing a person walking a large fluffy dog through a colorful, surreal cityscape with the title “Hard Problems Are Still Hard.”

Hard problems are still hard: A story about the tools that change and the work that doesn't

Figma designer–turned–product manager Nikolas Klein worked on building prototyping tools for seven years. Then AI changed the game.

. People of all ages are turning to LLMs for everything from relationship advice, to financial guidance, to home project planning, to research. In a world where computers understand intent as well as instruction
Abstract digital illustration of a hand releasing pixelated butterflies and data clustersAbstract digital illustration of a hand releasing pixelated butterflies and data clusters

5 shifts redefining design systems in the AI era

As AI reshapes how we make products, design systems are evolving from libraries of reusable parts into living frameworks that scale taste and craft. We spoke with product leaders and practitioners about the shifts they’re seeing in how design systems are built, used, and maintained.

, more people can make more things with fewer resources, whether that’s an app to help local farms
a yellow numeral three overlaid on a dark green numeral threea yellow numeral three overlaid on a dark green numeral three

Version control: One founder’s mission to save local farms with Figma Make

Factory farms and rising costs are putting British Columbia’s small growers out of business at an alarming rate. To fight back, entrepreneur Aaron Veale used Figma Make to build a marketplace app that helps farmers sell their produce—all in a matter of weeks.

 or a web poetry project

6 winning Figma Makes—and what you can learn from them

Winners from our first global Make-a-thon offer insights on how to prototype smarter, structure products better, and push Figma Make further.

. Natural language prompts make software feel organic and technology feel accessible in unprecedented ways, but they’re still bound to a chat interface. In a future where we communicate with machines through other means, we’ll open up new ways to translate human vision into actual impact
Four cartoon chefs gather around a glowing pink cake with cherries as one prepares to slice it.Four cartoon chefs gather around a glowing pink cake with cherries as one prepares to slice it.

How to harness skills that AI can’t automate

With more teams moving faster than ever, the real differentiator is craft—the curiosity, intuition, taste, and intention behind every detail.

.