bug-bounty524
xss282
rce152
bragging-post120
google112
account-takeover99
exploit96
open-source91
microsoft81
csrf78
facebook77
authentication-bypass75
stored-xss72
privilege-escalation72
cve70
access-control65
ai-agents63
malware63
reflected-xss61
writeup53
input-validation53
web-security53
ssrf52
sql-injection49
cross-site-scripting48
smart-contract46
tool46
defi45
ethereum45
privacy44
web-application43
phishing40
information-disclosure39
apple38
browser38
responsible-disclosure37
llm37
web337
lfi36
burp-suite35
opinion35
api-security35
cloudflare34
vulnerability-disclosure34
automation34
oauth33
idor32
machine-learning32
dos31
code-generation31
0
2/10
A longitudinal study analyzing 40 companies from November 2024 through February 2026 found that despite 65% increase in AI adoption, pull request throughput only increased by ~10%, contradicting vendor marketing claims of 2-3x productivity gains. The research suggests coding was never the bottleneck, and non-coding activities like planning, review, and alignment remain untouched by AI.
ai-productivity
developer-productivity
research
metrics
longitudinal-study
github-copilot
engineering-management
Justin Reock
DX
Abi Noda
Engineering Enablement
GitHub Copilot
Cursor AI
METR
Amplitude
Figma
Plaid
UserTesting
Vercel