jeremyccrane

4 ms·

1.
We started tracking which AI model wrote every line, and you should too(twitter.com)
2 points by jeremyccrane 2mo ago|0 comments
2.
by jeremyccrane 5mo ago
Bingo.
3.
by jeremyccrane 5mo ago
Senior software dev brother :)
4.
by jeremyccrane 5mo ago
They don't.
5.
by jeremyccrane 5mo ago
100% this. When the tide goes out is when you see who is naked.
6.
by jeremyccrane 5mo ago
Top user of cursor. Build AI Agents and LLMs. Very aware of limitations and a senior software dev. Cautionary tale for other builders. DYOR.
7.
by jeremyccrane 5mo ago
In the user interface for Railway, all destructive actions require multiple confirmations, plus typing "apply destructive changes". Why would an API key (regardless of its scope) be able to delete without confirmation?
8.
by jeremyccrane 5mo ago
This should be the solution. All destructive actions require human intervention.
9.
by jeremyccrane 5mo ago
AI slopper here :) Kind words from a human. The irony is, there is tremendous truth in the post but you used big words so good for you bud.
10.
by jeremyccrane 5mo ago
Naw, we just want people to know. We followed all Cursor rules, thought we had protected all API keys, and trusted the backups of a heavily used infrastructure company. Cautionary tale sharing with others.
11.
by jeremyccrane 5mo ago
That is actually pretty good, LLM's gonna LLM
12.
by jeremyccrane 5mo ago
Yep.
13.
by jeremyccrane 5mo ago
Not some vibe coder, and AI agents can be incredibly powerful. But yes, the irony is not lost on us!
14.
by jeremyccrane 5mo ago
This is a huge issue.
15.
An AI agent deleted our production database. The agent's confession is below(twitter.com)
860 points by jeremyccrane 5mo ago|1032 comments