Reply
· · to read

In reply to @KaraLG84.

@KaraLG84 @talon I wish more people would really think about this. Nobody raising the alarm about the current AI hype is suggesting that large language models and other machine learning tech can't be useful. What we're saying is that these are not silver bullets, and the potential harms aren't that they'll become sentient and take over the world. They're real-world happenings that effect actual people right now.

Repost
· · to read

I should probably post this here too for good measure, since I think this is very, very important.

We’re all happy that we finally have AI to help us describe images, but please, please keep in mind that this tech is the same tech that can’t tell you accurate information using ChatGPT. This thing will lie to you. It will happily tell you incorrect information, and it won’t know it did it. This is not a magic bullet! Please be careful! Do not trust it with your life, for example in regards to medication, or other high risk situations! Please keep this in mind!

—@talon