Lightnews — Scholar-powered news

LightNews

IT Revolution

@itrevolution.com

270 followers 470 following 710 posts

Helping technology leaders succeed through books (The Phoenix Project, Team Topologies), events (DevOps Enterprise Summit), research, podcasts (The Idealcast), and more.

Posts Media Videos Starter Packs

IT Revolution @itrevolution.com · 11h

Coding agents see their own mistakes if you let them. Yegge and Kim explain that they can automatically access browser console, terminal, logs, test suites. Steve asked: "Resurrect all deleted tests between 20-100 commits ago." Claude Code did the Git surgery. This is real today.

itrev.io/4mKG15N

IT Revolution @itrevolution.com · 14h

Kim's Fifth Ideal—Customer Focus—is now the foundation of product-led growth. Platform engineering embodies this by treating internal developers as customers. "Platform as product" mindset thinks about developer goals, not just tasks. The Unicorn Project predicted this shift.

itrev.io/474GafM

IT Revolution @itrevolution.com · 14h

Your first AI prompt is longest. After that? "Yes, go!" "Explain #2 further." "Use conventions in create-drafts-and-rank." Kim and Yegge show this pattern in Vibe Coding. When AI earns trust, prompts get shorter. The goal: shipping code, not wordsmithing prompts.

itrev.io/4mKG15N

IT Revolution @itrevolution.com · 16h

Devs spend 70% of time on business tasks, maintenance, and legacy infrastructure. Only 30% on actual coding. This fragmentation validates Kim's Second Ideal: Focus, Flow, and Joy.

Protecting flow isn't luxury—it's the only way to deliver value. The Unicorn Project shows how.

itrev.io/474GafM

IT Revolution @itrevolution.com · 18h

Vague requests get vague results. Kim and Yegge explain: "We need to handle dates with time zones" gives AI nothing. But messy dictation about what you know and what you want? AI figures it out like an attentive person would. You need clear problems, not perfect grammar.

itrev.io/4mKG15N

IT Revolution @itrevolution.com · 1d

AI has read almost everything on the internet. Steve Yegge and Gene Kim show how to leverage this in Vibe Coding. Working with ffmpeg? "Extract a 30-second clip starting at 2:15, remove audio, compress to 720p." Stop learning cryptic tooling. Start solving problems.

itrev.io/4mKG15N

IT Revolution @itrevolution.com · 1d

Despite heavy AI investment in 2024, developers still aren't seeing marginal productivity gains. Adding more tools doesn't solve complexity. Kim's emphasis on systemic thinking over point solutions explains why. The Unicorn Project is the antidote to tool-based thinking.

itrev.io/474GafM

IT Revolution @itrevolution.com · 1d

Platform engineering's J-curve: boosts individual productivity 8% and team performance 10%, but initially decreases throughput 8% and stability 14%. Kim depicted this transformation challenge perfectly. Organizations that expect instant wins abandon efforts. The prepared succeed.

itrev.io/474GafM

IT Revolution @itrevolution.com · 1d

Most powerful debugging technique for AI? Copy and paste the error message. Kim and Yegge explain that AIs are remarkably good at understanding errors. Instead of describing what broke, show the stack trace. Let errors speak for themselves. The visual info is all AI needs.

itrev.io/4mKG15N

IT Revolution @itrevolution.com · 1d

Stop crafting perfect prompts. Gene Kim and Steve Yegge's book Vibe Coding shows why treating AI like texting—not a legal brief—gets better results. When AI makes mistakes, refine and redirect. Your typos don't matter. Getting to working code matters.

itrev.io/4mKG15N

IT Revolution @itrevolution.com · 1d

Kim's Second Ideal—Focus, Flow, and Joy—moved from fiction to the boardroom. Half of tech managers now measure developer productivity or DevEx. 16% have dedicated productivity specialists.

This isn't soft skills. It's strategic.

And The Unicorn Project predicted it.

itrev.io/474GafM

IT Revolution @itrevolution.com · 2d

Bill Palmer thought he was making progress. Then the Phoenix deployment spiraled into disaster. The Phoenix Project Graphic Novel Vol 2 raises the stakes: 90 days to transform or face outsourcing.

For anyone leading or living through transformation.

itrev.io/4glAS0S

IT Revolution @itrevolution.com · 2d

Top dev productivity blocker in 2025: time to gather context (31% cite this).

Gene Kim's Third Ideal—Improvement of Daily Work—is now the foundation of platform engineering. The shift from "ship features" to "improve how we ship" is real. The Unicorn Project showed us how.

itrev.io/474GafM

IT Revolution @itrevolution.com · 2d

High psychological safety = 50% more productivity, 76% more engagement, 27% less turnover. Yet only 50% of workers say managers create it. Gene Kim's Fourth Ideal predicted this. For complex systems, psychological safety enables the innovation you need.

Now in paperback.
itrev.io/474GafM

IT Revolution @itrevolution.com · 2d

Kim's First Ideal—Locality and Simplicity—reads like a warning we ignored. We rushed to microservices without considering cognitive load. Now we have systems too complex to understand, debugging like archaeology.

In 2025, The Unicorn Project is required reading.

itrev.io/474GafM

IT Revolution @itrevolution.com · 3d

The AI paradox validates Gene Kim's systemic thinking: 75% of devs report productivity gains from AI, yet DORA shows AI decreases delivery throughput by 1.5% and stability by 7.2%.

Individual productivity ≠ system improvement. The Unicorn Project explained this before AI existed.

itrev.io/474GafM

IT Revolution @itrevolution.com · 3d

Platform engineering barely existed when The Unicorn Project published. Now 55.84% of platform teams are less than two years old. Platform engineers earn 26.6% more than DevOps roles. Gartner predicts 80% of software orgs will have platform teams by 2026.

Gene Kim called it.

itrev.io/474GafM

IT Revolution @itrevolution.com · 3d

Gene Kim's The Unicorn Project showed a developer drowning in cognitive overload. It felt like fiction. Today, 72% of orgs report new hires take over a month for their first meaningful PRs. Developers spend only 30% of time coding. Kim wrote a documentary of 2025, five years early.

itrev.io/474GafM

IT Revolution @itrevolution.com · 4d

Your users aren't lazy, change-averse, or technologically backward. They're intelligent people managing complex workflows where stability matters. They're making rational decisions with limited change capacity.

Build delivery systems that work with them, not against them.

itrev.io/3IQlyOW

IT Revolution @itrevolution.com · 4d

Netflix tests thousands of interface variations but users never feel stuck because changes are subtle or easily reversible. When people trust they can explore without getting trapped, they're more willing to try changes. User confidence enables innovation.

itrev.io/3IQlyOW

IT Revolution @itrevolution.com · 4d

Yassir scaled from 50 to 400 engineers with 140% increase in deployments and 230% boost in employee satisfaction using their "SAUCE" approach: Small, Audience focused, Cross-functional, Enabled teams. That's how you scale right.

itrev.io/423AMqq

IT Revolution @itrevolution.com · 4d

You can't build empowered teams by copying org charts. Skelton and Pais's Team Topologies 2nd Edition gives you four team types, three interaction patterns, and one critical insight: team cognitive load is your key design constraint.

itrev.io/4nTtkq8

IT Revolution @itrevolution.com · 5d

GitHub ships hundreds of small changes that are mostly invisible to casual users but meaningful to developers spending all day in the platform.

Meanwhile, major releases are carefully communicated and gradually rolled out.

Multiple speeds for multiple constituencies.

itrev.io/3IQlyOW

IT Revolution @itrevolution.com · 5d

Microsoft's "Try the new Outlook" toggle. Google Workspace's separate release tracks. These aren't just beta flags—they're ongoing control over experience.

That's the difference between respecting change capacity and forcing adoption timelines.

itrev.io/3IQlyOW

IT Revolution @itrevolution.com · 6d

Three strategies:

1) Delegate control to users—let them decide when to adopt features.

2) Design for multiple speeds simultaneously—power users and cautious users don't need the same pace.

3) Build reversible experiences—make experiments safe by making them undoable.

itrev.io/3IQlyOW