Tech News
← Home  ·  All topics

Figma

9 GoKawiil briefs on this topic

Figma trains custom AI model to edit designs directly on canvas

Figma is developing an in-house editing agent trained via supervised fine-tuning to understand its native file format and make iterative design changes, such as spacing, color, or layout edits, directly on the canvas. The company built training data by pairing starting designs with synthetic edit requests and validated outputs, using larger teacher models, deterministic validators, and visual judges to check results. Figma reports that an early Llama 3.3 70B model trained with this approach performed well on these tasks.

Apple adds touchscreen MacBook Pro assets to its Design Resources portal

Apple has updated its official Design Resources portal with Photoshop and PNG assets for the newly announced MacBook Pro featuring a PROLED XDR Display and touchscreen support. The downloadable files include transparent PNG images in multiple colors and sizes, plus PSD files with masked display areas for designers to insert their own interface mockups.

Figma traces decades-old Shift+Numpad bug to IBM's 1977 keyboard design

A Figma engineer investigated a user-reported quirk where Shift+2 worked differently depending on whether it was pressed via the main keyboard or the numeric keypad, but only on Windows. The investigation traced the behavior back to keyboard layout decisions IBM made starting with its 1977 System/34 and carried forward through the Datamaster and the original IBM PC.

Figma adds 20 MCP partners while Field voices doubts about the protocol

Figma announced 20 new model context protocol (MCP) partners at its first Config conference in India, including Amp, ClickUp, Dia, Junie by JetBrains, Linear, Meta Muse, Mistral, Notion, OpenCode, and Sarvam Agents. The expansion comes even as co-founder Dylan Field has publicly questioned MCP's design, previously criticizing Anthropic for not exposing its own resources via the standard, and he has said Figma plans to support the open-source tool Pi.

Figma launches Specs editor and expands code layers to guide AI agents

At its first Config event in India, Figma introduced Figma Specs, a Markdown-based editor that lets teams define design language and guidelines as a reference for both humans and AI agents. The company also moved its code layers feature, which adds code to canvas designs for interactive prototypes, from private to public beta, and announced integration with the Payload content management system.

Figma limits remote MCP server access to approved clients, excludes Pi

A Figma representative confirmed that the company's remote MCP server only works with clients on an approved allowlist, and the Pi client is not currently included. Figma pointed to its public MCP catalog listing supported clients and directed developers to a form to request that Pi be added in the future.

Motif Design brings complex architectural building models to browser-based collaboration

Motif Design has launched a browser-based platform that lets architecture teams work with highly complex building models directly online, without specialized desktop software. The tool is positioned as a collaboration-first alternative to traditional architectural modeling workflows, echoing the browser-native approach that made Figma popular in design.

Pause launches weekly curated 1:1 coffee matching service for London professionals

A new service called Pause pairs London professionals for weekly 45-minute one-on-one coffee conversations based on neighbourhood proximity, aiming to strip away small talk and agendas. The launch, shared as a Show HN post, also outlines related offerings including skill-swap matches between people with complementary expertise and a curated list of quiet cowork spots in independent cafes.

OpenAI expands ChatGPT Work to bring agentic AI to non-engineers

OpenAI has launched ChatGPT Work, a $20/month product adapted from its Codex coding tool, aimed at giving white-collar workers—not just software engineers—access to AI agents that can autonomously complete multistep tasks. To function effectively, the tool requires deep access to users' digital lives, including email, Slack, phone data and apps like Notion and Figma, as demonstrated by OpenAI's own lead engineer who granted it broad control over his accounts.