Working well with coding agents takes months, not hours
Experienced engineers resist agentic coding, and most people quit after a couple of hours. The value only shows once you have put in the time to learn the new skills.
Written by
-
Benedikt Stemmildt Founder & Co-CEO
Published: 29 January 2026 Updated: 2 min read
Senior engineers hate vibe coding. Junior engineers are outshipping them.
Steve Yegge said this at the AI Engineer Summit: the group most resistant to agentic coding is not the juniors. It is engineers with 12 to 15 years of experience. Their identity is tied to the old ways.
The performance gap is already visible
Yegge describes internal data at OpenAI showing large performance gaps. Same title, same job. One engineer uses agentic loops, the other does not. At performance review time, HR is asking legal what their options are.
This is no longer a debate about code quality. It is about survival.
The 2,000-hour rule
Yegge and Kim put it at a full year before you can predict what the agent will do. Before that, it feels random, frustrating and untrustworthy.
Most engineers quit after two hours. They conclude it is garbage and move on. The ones who pushed through cannot imagine going back.
What vibe coding gives you: FAAFO
Gene Kim and Steve Yegge describe the value in five parts:
- Faster. You build things much faster. Ideas that sat on a “someday” list for years get done in a weekend.
- Ambitious. When execution costs drop, your threshold for “worth building” drops too.
- Alone. Things that required a team are now possible solo. The bottleneck shifts to whether you can think clearly about what you want.
- Fun. When the tedious parts disappear, the creative parts are left: the architecture decisions.
- Optionality. More swings at bat. Fail faster, learn faster, explore more before committing.
You cannot skip the learning
You can’t grab the Batman suit and become Batman. You’re just cosplaying.
Working this way requires new skills. Reading the shape of a diff. Noticing when the agent writes suspiciously much code. Knowing when to trust and when to verify.
This takes months, not hours.
From the IDE to an orchestration dashboard
The IDE as we know it is dying. What replaces it is an agent orchestration dashboard rather than another editor. You walk in, check which agents need input, review their work, and move on.
We’re moving from cutting corn with scythes to factory farming code. Steve Yegge and Gene Kim make the full case in their talk on vibe coding (opens in a new tab).
Are you investing in these skills, or betting your career that this is all hype?
Further reading
Written by
-
Benedikt Stemmildt · Founder & Co-CEO
More on Adoption
- AI makes teams faster and wears down their engineering instincts Agents speed up output while problem decomposition, debugging and code comprehension fade. The answer is to encode the team's engineering knowledge into a harness.
- Three windows for adopting agentic coding are closing Teams that plan to adopt agentic coding in two or three years underestimate the learning curve, the talent market and the economics. Each of the three gives a reason to start the clock now.
- Five human skills that grow more valuable with AI My son asked why he should learn programming if AI will do it for him. The skills that matter most are becoming more valuable: critical thinking, empathy, language, value creation, judgment and taste.
Test the constraint in your system
The Factory Readiness Assessment connects this research to your value stream, your data, and the next decision.
Training on this topic: Agentic Engineering for Engineering and Product LeadersWant to check one team first? The self-check: twelve yes or no questions, three minutes, instant result, no account. Take the self-check