🎮UX in AI #day6 • When implementing AI in your product, define a clear pattern. • A pattern that user's will recognize every time they want to interact with your feature. • Recognition that comes when widely distributed across your product. See this example by Notion.

(bad) UX in AI One of the flaws of AI on X. 150 of my followers are made-up bots. This means jt looks like you have a lot of credibility because of your follower base. But in fact, you don’t.

🎮 UX in AI #day8 Designing for AI isn’t just designing components. It’s also: • Define user needs • Narrate stories • Try and iterate • Look around to user responses • Handle errors • Not overpromise Equally important if you want successful UX design

Nah. UX is still the same. AI is creating just another paradigm to find opportunities for user or define user needs.

🎮 UX in AI #day8 Designing for AI isn’t just designing components. It’s also: • Define new user needs • Narrate stories • Try and iterate • Look around to user responses • Handle errors • Not overpromise Equally important if you want successful UX design

🎮 UX in AI #day9 • OpenAI new feature is called Canvas • A new pattern introduced is to highlight specific areas, let OA suggest edits, and rewrite. • This gives you as a user control with the system's support.

Great cheat sheet by Matt Przegietka. Would this work with products that are AI-powered?

🎮UX in AI #day10 The problem with Canvas? Unlike @Grammarly or Cursor, these are standalone tools that work in current workflows. Write in Google Docs and code in VS Code. Canvas is positioned as if you code in ChatGPT and write documents in ChatGPT. (Credits Ethan Mollick)

🎮UX in AI #day11 • Don't trust AI in the first place, so you can more trust in the future. • Why? By being critical and error-friendly to the output from Gen AI, your user experience will be better. • And UX is about trust.

🎮UX in AI #day12 • Strava just released Athlete Intelligence. • Encouraging, analytical, and insightful. • However, with just text, it's hard to analyze. And one of the strengths of Strava is their detailed analysis.

🎮UX in AI #day14 • 95% of user of your product don’t care about AI. • Why? • They care about how you can help them and you care about low costs. • So, question: Can I do this well enough without AI?

Back since like a year or so. Never really wanted to be on X anymore, but DesignTwitter is dragging you back in. Had an absolute blast using MagicPath I build a tool called Podsight, that can analyse and parse podcast for resources and insights. designs.magicpath.ai/v1/happy-river…