Mark Levison on Nostr: 📊 "Did the code pass the test?" isn't enough. Current AI coding benchmarks miss ...
📊 "Did the code pass the test?" isn't enough.
Current AI coding benchmarks miss crucial factors: readability, maintainability, fitting existing patterns.
Senior developers in study were actually slower with AI despite feeling faster. We need objective measures, not just developer surveys.
https://agilepainrelief.com/blog/using-genai-to-code-not-so-fast/ 
Published at
2025-08-19 14:08:19 UTCEvent JSON
{
"id": "eff4c758fb42f4979a6f4599036486c5c2f84fc05a58f1b6df65c5b180fc406d",
"pubkey": "8985a9535e310be58f29ce043a421773c9396d4bbbc048f9c637665e11d98c1f",
"created_at": 1755612499,
"kind": 1,
"tags": [
[
"imeta",
"url https://agilealliance.social/system/media_attachments/files/115/055/820/712/386/616/original/bbbb7f08de605ec7.jpeg",
"m image/jpeg",
"dim 800x600",
"blurhash U2SijYIVxu%20f0KIU~W?c0eIV~W.8AG?b^%"
],
[
"proxy",
"https://agilealliance.social/users/mlevison/statuses/115055820782817408",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.mostr.pub"
]
],
"content": "📊 \"Did the code pass the test?\" isn't enough.\nCurrent AI coding benchmarks miss crucial factors: readability, maintainability, fitting existing patterns.\nSenior developers in study were actually slower with AI despite feeling faster. We need objective measures, not just developer surveys.\nhttps://agilepainrelief.com/blog/using-genai-to-code-not-so-fast/\n\nhttps://agilealliance.social/system/media_attachments/files/115/055/820/712/386/616/original/bbbb7f08de605ec7.jpeg",
"sig": "5396dface08dad7aab6794b201c4ad4a12c02c9ef8e12fbb3c613d3eb927b36b497456731631abbcba11324776b73b499102682b816f013d3b70bcf5ec94f5e1"
}
