2 months ago
- Competitive programming faces a new era with AI, where solutions can be easily generated, leading to widespread cheating.
- AI-assisted unfair players cause rating deflation, estimated at 150-190 points, harming genuine participants and creating a vicious cycle.
- Recommendations for genuine contestants include detaching from rating, participating in offline contests, and using virtual contests from pre-AI eras.
- Using AI for practice should be limited; avoid relying on it for debugging to develop critical skills manually.
- Proposed contest structure changes include AI-based cheating detection, rating performance monitoring, public reporting, and stricter account verification.
- Despite AI challenges, competitive programming retains value in training problem-solving skills, with a focus on personal growth over leaderboard rankings.