Slopfix is a team of three senior engineers who refactor vibecoded codebases back to maintainability.

No, Slopfix isn't a sponsor 😉

Why Your 360 Feedback Isn't Working

• 🧠 The perception gap — Genuine behavior change often fails because others’ mental models of you are rigid and slow to update due to confirmation bias.
• 👥 Selective engagement — Focus your growth efforts on the small group of people who provided the feedback or those you most need to influence, rather than a broad, impersonal survey group.
• 📂 Feedback triaging — Categorize feedback into three buckets: "Sign it" (commit to change), "Realign it" (translate to your authentic style), or "Decline it" (ignore if it conflicts with core values).
• ⚙️ Action over concepts — Replace vague development goals like "be more strategic" with specific, observable, and trackable actions that can be scheduled into your routine.
• 🗣️ Explicit collaboration — Tell feedback providers exactly what you are working on and ask for their help in monitoring your progress to ensure they notice when you actually change.
• 📊 Systems vs. motivation — Do not rely on fleeting willpower; design a simple, recurring system—such as weekly self-scoring—to objectively track your behavior independently of your mood.

Ikigai

• 🎯 The core concept — Ikigai is a Japanese term meaning "a reason for being," representing the intersection of what you love, what you are good at, what the world needs, and what you can be paid for.
• 🧘 Internal fulfillment — The term is deeply rooted in Japanese culture as a personal, subjective state of wellbeing rather than a strictly professional or external career pursuit.
• 📉 Cultural evolution — While originally used to describe one's source of meaning in everyday life, the concept has been Westernized and popularized as a Venn diagram framework for finding the perfect career.
• 🕰️ Longevity connection — The word gained global attention for its association with the residents of Okinawa, where it is cited as a significant factor in the community's exceptionally high life expectancy.
• ⚖️ Fluidity over rigidity — True ikigai does not require all four elements to be perfectly balanced at once; it is a dynamic pursuit that can shift based on an individual's life stage and circumstances.

Using AI to write better code more slowly

• 🎯 AI coding is most effective when used to write high-quality code slowly, not to rapidly churn out low-quality "slop" — the tools are flexible enough to support either approach.
• 🔍 LLMs excel at finding bugs in codebases when deployed repeatedly, but the real challenge is prioritizing and validating findings rather than discovering them.
• 🤖 Running multiple different models on pull reviews (Claude, Codex, Cursor Bugbot) dramatically reduces hallucinations and false positives — near-zero false positive rates are achievable.
• 📋 A practical workflow involves: have agents fix critical/high-priority bugs iteratively, skip medium-level issues where effort isn't justified, and abandon PRs if they reveal fundamental architectural flaws.
• 📈 This slower approach often doesn't increase velocity but surfaces pre-existing bugs, leading to tangential "side quests" writing tests and fixing subtle flaws — which improves overall codebase health.
• 🧠 Understanding failure modes and edge cases of complex systems is more valuable than the happy path — this mirrors how developers learned codebases before LLMs existed.
• 🎓 The technique doubles as excellent onboarding for unfamiliar codebases, helping build mental models faster than reading documentation.
• ⚠️ Junior developers need significant domain knowledge to triage AI findings — agents find bugs, but developers must understand them to distinguish real issues from theoretical ones.
• 💭 AI doesn't change behavior patterns; it amplifies them — the system makes unreviewed deployment tantalizingly easy, requiring deliberate discipline to work methodically.

What Most People Miss About Performance Reviews

• 📊 Manager approval isn't enough — At senior levels, peer perception and organizational reputation become equally or more important than your direct manager's assessment for securing promotions.
• 🏢 Calibration sessions reshape your narrative — Performance reviews expand from your immediate team to "calibration sessions" where dozens or hundreds of leaders discuss and compare your performance across functions, making your broader brand critical.
• 😲 Surprises happen in the room — Promotion candidates can face unexpected objections from peers they've never heard from, including concerns about collaboration, credit attribution, or metric accuracy that derail promotions.
• 🎯 Proactive relationship building is essential — Months before promotion conversations, you and your manager should seek feedback from key influencers across the organization to identify potential advocates and address concerns early.
• 🤝 Find independent advocates — Recruit influential leaders outside your direct reporting line to mentor you and vouch for your capabilities, since your manager may be perceived as biased.
• 🗣️ Control the narrative — Ensure others are "buying into" your wins and impact story; make sure your accomplishments align with organizational priorities and are understood broadly.
• ⚠️ Reputational risk at smaller companies — Even without formal reviews, smaller organizations have more casual performance discussions where influence is obvious and reputation spreads faster.

A return to two-pizza culture

• 🍕 Two-pizza team culture remains foundational — Small teams where everyone knows what peers are working on, with autonomy to make reversible decisions without bureaucracy or permission
• 📝 Working Backwards process traditionally started with writing — Press releases, FAQs, customer experience docs, and user manuals forced clarity before building, and democratized product definition across non-engineers
• ⚡ AI coding agents are compressing the prototype phase — What used to take months now takes an evening, shifting the question from "should we build this?" to "does this feel right to use?"
• 🔄 The process needs amendment, not replacement — Building prototypes first (when uncertain about approach), then using them like customers, then writing docs grounded in reality produces better outcomes than writing-first
• 👥 Amazon Quick desktop exemplifies the new model — One engineer built a prototype overnight with an AI agent; the team grew organically as people saw it and wanted to contribute, operating with full autonomy
• 🎯 Experience-first beats benchmarking-first — Team members used the product daily from day one, fixing rough edges immediately rather than deferring UX work; code reviews included videos of the actual experience
• 📊 Growth doesn't require abandoning the culture — Quick scaled to hundreds of people while maintaining two-pizza team structure through clear ownership, autonomy, and short feedback loops
• 💡 The core insight: "You will learn more in one evening of building than in two weeks of writing" — Prototyping now compresses learning cycles and pressure-tests assumptions in ways documentation alone cannot

That’s all for this week’s edition

I hope you liked it, and you’ve learned something — if you did, don’t forget to give a thumbs-up, add your thoughts as comments, and share this issue with your friends and network.

See you all next week 👋

Oh, and if someone forwarded this email to you, sign up if you found it useful 👇