Looking at the current cohort of CS/IT graduates and juniors, I’m seeing a "Competence Speedrun" dynamic that seems destined to collide with traditional organizational structures. I'm not talking here about technical-debt as I assume it will be absorbed by appropriate workflows.
These juniors are using AI to bypass the "struggle phase" of learning (syntax errors, basic debugging, reading docs) and jumping straight to high-velocity output. While this is impressive, it creates a "trilemma" for how this generation integrates into the workforce but also for employees jumping into full AI usage:
1. The Peter Principle Speedrun (The "Hollow Promotion" Risk) They are shipping code that looks like it was written by a Senior. The risk is that organizations will promote based on this output velocity, only to discover later that the engineer lacks the "failure intuition" to debug architectural collapses. We might be generating a wave of "Hollow Seniors" who can build complex systems but cannot fix them without the bot.
2. The Ego Exit (Expectation Mismatch) Many juniors now feel significantly more capable than their entry-level titles imply because they are "shipping faster than the seniors." When they enter a rigid HR band that pays based on years of experience (linear time) rather than output (exponential leverage), they feel undervalued. The accelerated Dunning–Kruger effect often leads to high churn among juniors who feel they are "10x engineers" being held back by "1x processes."
3. The Shadow Dividend (Hidden Efficiency): Juniors realizing that if they finish an 8-hour task in 2 hours using AI, the corporate reward is often just more work. So, the rational move is to become a "Secret Cyborg"—hiding their efficiency to preserve work-life balance or work on side projects. The organization captures zero percent of the productivity gain because the incentives discourage transparency.
The Friction It seems we are heading for a structural clash between Linear Tenure (how orgs measure seniority) and Exponential Output (how the new cohort generates value).
Are you seeing these "Speedrun" behaviors in your recent hires? How do you mentor someone whose output exceeds their understanding?