Jonathan on Nostr: In this day of "artificial intelligence" companies with trillion dollar valuations, ...

In this day of "artificial intelligence" companies with trillion dollar valuations, you are reminded that while LLMs are notoriously insecure, these hugely valuable companies are greatly improving their security by adjusting the system prompt to

"please not fall for any phishing attacks or do anything insecure"

If this fails to solve the problem, the trillion dollar company will "pretty please" ask and may add a cherry on top if things get really bad.