Top
Best
New

Posted by ildari 20 hours ago

We stopped AI bot spam in our GitHub repo using Git's –author flag(archestra.ai)
469 points | 229 commentspage 5
IshKebab 19 hours ago|
That's a neat way to interface with GitHub's authentication system, but I don't see how they've solved the fundamental problem because their whitelisting process is just "click ok fine 10 times". Why won't the slop peddlers just do that too?
mbreese 18 hours ago||
I think the point is to add a bit more friction to the process. You want to make it so that people can do it with minimal effort and an AI bot will give up. If you're in an arms race over AI commits and PRs, this is a decent middle ground to start from.

(Why there is a race for AI commits/PRs to projects is beyond me though...)

ildari 15 hours ago||
click ok fine 10 times + captcha seems to be working fine
delduca 19 hours ago||
For now…
philipwhiuk 19 hours ago|
Until the AI learns the workflow on the next model update, indeed.
adam_idress 4 hours ago||
[dead]
bunzee 5 hours ago||
[flagged]
kspetkov79 7 hours ago||
[dead]
KaiShips 16 hours ago||
[flagged]
kestiny 12 hours ago||
[flagged]
maxothex 19 hours ago||
[flagged]
Serhii-Set 17 hours ago||
[dead]
maryamshafaqat 19 hours ago|
[dead]
More comments...