3 ms·
Every time someone has used an automatically publish feature of any type, it always ends poorly. Either personal information gets posted they didn't intend, inc
by gochi 3y ago
Every time someone has used an automatically publish feature of any type, it always ends poorly. Either personal information gets posted they didn't intend, incomprehensible grammatical errors, or as in this case completely falsified information gets released.
It's a very old lesson we should have learned from newspaper days but I guess with how fast people in AI are moving they don't care about "old lessons".
Outside of that, it's interesting to see how people try to combat and highlight AI platforms. Jokes like these, paywalls, limited invites, increasing API costs, and so on. Very interesting times for online information.
- pixl97 3y agoOld paradigm: A lie can travel halfway around the world before the truth can get its boots on New paradigm: An AI can conquer half the world before the killswitch operator can pick their axe up.
- lapetitejort 3y agoThat's why it's important to suspend the axe above the wires ahead of time. It's also important to to remind the AI that the axe exists. The Axe of dAImocles
- chiph 3y agoAbout ten years ago there was an Agile conference that published a hashtag where anything that used that tag would automatically get shown on the monitors around the building. 4chan found out, and well, things went predictably. Automatic publishing is always going to be vulnerable to content poisoning.