#Anthropic calls it the Supervision Paradox:
Overuse AI → Lose coding skills → Can’t verify AI.
The irony? You need the skills AI erodes to supervise it.
Who will review AI-generated code in 5 years?
🔗 Watch the full video: buff.ly/w0lDShK
#SoftwareEngineering #AI #Coding #InfoQ
#Anthropic calls it the Supervision Paradox: Overuse AI → Lose coding skills → Can’t verify AI. The irony? You need the skills AI erodes to supervise it. Who will review AI-generated code in 5 years? 🔗 Watch the full video: buff.ly/w0lDShK #SoftwareEngineering #AI #Coding #InfoQ
0 vue
Publié le 05/11/2026 11:48
0:38
#Anthropic calls it the Supervision Paradox:
Overuse AI → Lose coding skills → Can’t verify AI.
The irony? You need the skills AI erodes to supervise it.
Who will review AI-generated code in 5 years?
🔗 Watch the full video: buff.ly/w0lDShK
#SoftwareEngineering #AI #Coding #InfoQ
1 commentaire
?
Part of the answer is making review less dependent on memory. Tests, SAST and dependency checks enforced in CI catch the mechanical stuff. Humans then review design and intent. Juniors still need to write and debug unaided sometimes, or the reviewer pipeline dries up.