866: 2025 Web Development Predictions

Episode Summary

Executive Summary: Wes and Scott forecasted major web-development trends for the year: broader browser support for Temporal, more on-device AI powered by WebGPU, a plateau in model capability but better AI workflows, wider adoption of web components/Web Awesome, a shift toward standards-based “vanilla” browser APIs, and continued momentum in CSS features, React tooling, Bun, Svelte, Laravel, and AI browsers. The episode mixes predictions with reflections on current support, tooling fatigue, and the growing importance of browser primitives.

Main Topics: Temporal and date/time improvements: Both hosts predict Temporal will ship broadly in browsers, calling it a transformative replacement for fragile date handling and praising its polyfill and API design. On-device AI and WebGPU: They expect far more AI running locally on user devices as model sizes shrink and WebGPU improves performance, especially for OCR, transcription, and other browser-native tasks. AI tooling plateau and workflow evolution: They argue model quality may plateau, but products will improve by combining current models better; text-to-UI, code assistants, and AI-assisted browser experiences will keep advancing. Web components and Web Awesome: They predict growing use of web components in reusable UI libraries, with Web Awesome serving as a high-quality example that could normalize standards-based component adoption. Standards-based web stack comeback: They foresee developers increasingly choosing browser standards first—HTML, CSS, fetch, streams, ESM, and web APIs—over framework-specific abstractions. CSS feature expansion and simplification: They expect relative color, style queries, mixins/functions, and other modern CSS features to keep advancing, making vanilla CSS more powerful and reducing the need for utility frameworks. Framework and ecosystem predictions: Other forecasts include VS Code matching Cursor, React compiler dropping Babel, React Server Components and React Native gaining more traction, Bun moving into PaaS, Vite/Rolldown progress, and possible Laravel/Statamic CMS moves.

Key Arguments: Temporal is already mature enough via polyfill that browser shipping is mostly a matter of implementation, not design. WebGPU will make local AI much more practical by letting models use the user’s GPU instead of slower WASM paths. AI won’t necessarily get much smarter every year; instead, products will improve by orchestrating existing models better and by finding high-value workflows like text-to-UI and code generation. Web components will become easier to adopt through polished libraries like Web Awesome, encouraging standards-based UI reuse. Developers are increasingly fatigued by framework complexity and will lean more on browser primitives such as CSS, HTML, fetch, and ESM. CSS is becoming powerful enough that many common UI tasks no longer require heavy abstractions, and browser support is now good enough for wider adoption. AI-driven browsers and search experiences are strategically important because they may become the next primary internet entry point. Bun, Laravel, and other ecosystem players are likely to expand into adjacent products and platforms as they grow and seek monetization paths.

Data Points: Predicted browser hit rate: 82% - Wes jokes that their annual prediction accuracy is around 82%. Browser support for relative color: 88% support - Mentioned while discussing whether relative color is effectively ready for use. Historical support for relative color in Safari: almost two years - Used to argue the feature is sufficiently mature for adoption. Temporal browser availability target: end of the year - Prediction that Safari and Chrome will ship Temporal by year-end. WebGPU status: technology preview in Safari; behind a flag in Firefox; shipped in Chrome for a bit - Describes current browser support status for WebGPU. CSS mixins/functions status: intent to author - They say mixins/functions are being standardized but are not yet in browsers. Kids at home: 30 kids - Scott mentions dealing with strep among 30 kids at home.

Pivotal Quotes: "Temporal is absolutely massive and going to be such a game changer" — Wes: Explaining why the Temporal API is a major prediction for the year. "We’re going to see a play. Toe of AI" — Scott: Arguing that AI capabilities may plateau while products improve around the models. "The browser has this included. Maybe we should be using this because it's nice, it's fast, it's easy." — Scott: Supporting the prediction that developers will move toward standards-based APIs and away from framework-only solutions.

Implications: Listeners should expect more standards-first web development, richer CSS/HTML/browser APIs, stronger local AI experiences, and a continued shift in tooling toward AI-assisted workflows and reusable web components.