AI summary

• Mentorship is a high-impact career tool for designers at any level—it leads to more promotions, higher job satisfaction, better pay, and less burnout, not just early-career growth.
• Common reasons mentorships fail: choosing someone you can't be candid with, ignoring the mentor's advice, letting sessions become irregular, or picking a mentor without the relevant experience. Avoid your own manager or anyone who controls your career.
• Choose mentors 2–3 levels above you who have a track record of mentoring, and evaluate them early for credible, specific insight, candor, and vulnerability—reverse mentoring is also viable for emerging trends.
• Consider internal mentors for political and project insight, but external mentors with no ties to your employer often provide more honest, diverse perspectives. Use your network or formal programs like AIGA, UXPA, or Micromentor to find candidates.
• Reach out with a short, admiring email and propose a 30-minute meeting; don't self-reject. Even a senior leader you admire may be happy to mentor you—just remember that paid short-term coaching is not the same as ongoing mentorship.

AI summary

• The author built a design review skill for Claude, but discovered that listing eight design principles wasn't enough: a component could satisfy every principle yet still deliver a poor experience.
• The fix required three layers underneath the principles: precedence (which principle wins when they collide), prohibitions (outcomes never acceptable regardless of intent), and acceptance tests that check the rendered result rather than the intention.
• Most design systems only document the first layer, leaving the other three implicit — and the fourth is the one the author couldn't keep doing by hand.
• An acceptance test that depends on someone remembering to run it is just a checklist with good intentions; it needs to be automated or built into the review process itself.
• The hardest part of the project was teaching the skill to say 'I don't know' — a crucial behavior for honest design critique.

AI summary

• Every tech era introduces new UI idioms, and AI is already seeding its own—some will fade, but others may become as lasting as the hamburger menu.
• The sparkle emoji ✨ has been repurposed as shorthand for AI, often paired with rainbow colors—a quirky semantic shift that shows how quickly visual language evolves.
• AI-native interaction patterns like streaming text and shimmering 'thinking' placeholders are spreading beyond chat to signal any asynchronous task (loading, fetching, computing).
• Desktop AI apps (Claude, Codex, Cursor) share a distinct aesthetic of tiny icons, beige/cream backgrounds, orange accents, and serif typefaces—visually clashing with native macOS app conventions.
• AI's non-deterministic 'grain' also influences UI behavior, like whack-a-mole controls that repaint unexpectedly; designers should watch which aesthetics persist and which are just fads.

AI summary

• The Paradox of Choice means more information often leads to more anxiety and paralysis, not better decisions.
• Overthinking reduces working memory capacity, which directly impairs performance on mentally demanding tasks.
• A Stanford brain-imaging study found that the less participants deliberated, the more creative their drawings — overthinking kills creativity.
• Decision fatigue depletes willpower: even judges become more lenient after breaks, and agonizing over choices drains energy needed for later, higher-quality decisions.
• Maximizers who demand the best option are consistently less happy than satisficers who accept 'good enough' — a mindset shift can improve well-being.

AI summary

• Recognize AI fatigue when you start using AI for everything, including personal projects—set boundaries: use AI at work but not in your free time to preserve a sense of accomplishment.
• Do the fun parts yourself: hand-code CSS animations, write your own messages and articles; writing is thinking, and outsourcing it cheapens the process.
• Beware the '100x engineer' trap—chasing maximum productivity leads to burnout. Stay 10x more productive, but invest in work quality and satisfaction, not just output.
• You won't be left behind by skipping the newest AI tools; learn them only when essential. Example: skipping Sketch paid off when Figma became the standard, saving tool fatigue.