Do you think LLMs and other large broad models could/should be legislated to make them more regulated/predictable. The teenager committing suicide after speaking with chatgpt a lot is horrendous.
Replies
They will not become perfectly predictable with current tech. However they will likely do intense surveillance, individual risk scores, and notification to authorities of suspicious behavior. Pre-crime detection with a very high false-positive rate.
sep 1, 2025, 12:41 am • 0 0 • view
Does that require more urgent action than teens committing suicide after using social media, or with not tech involved?
aug 31, 2025, 8:34 pm • 0 0 • view
If someone spoke to a teenager and mentioned suicide thousands of times and told them how to make a suitable noise, then yes action should be taken.
aug 31, 2025, 8:43 pm • 1 0 • view