Prevent dangerous AI: Difference between revisions

Latest revision as of 01:15, 7 March 2026

Current dangers

Why it matters

1. Don't let governments become dictatorships. Footage is a key part of holding authorities accountable. Who will watch the watchers? We the people. We can't let AI ruin that.

Not being able to trust pictures, video, audio

FAQ

What about AI detection software?
Even though generators and detectors both get more advanced over time, the generators ultimately win, unfortunately. Read more

What could be done:

policy ideas

Ban unwatermarked generative AI

tech ideas

Anonymous camera digital signatures

Status quo:

Google's AI image generators produce images with a "synth ID" so they can be identified as fake. Not sure how foolproof it is though.

Sora AI (video generator) adds a subtle watermark to the video. Some people have been known to crop it out though.

There are plenty of image generators that don't produce any watermark at all.

Watermarks in general are not hard to remove, unfortunately. Still better than nothing though; most people probably won't bother removing them. So watermarking can at least reduce the total amount of misleading slop on the internet.

FAQ

Why not just create more jobs? Read more

Other concerns

* Environmental impact of AI
* AI psychosis

Other discussions

* Talk: Are AI images art or plagiarism?

Unemployment

Main solution: Socialize basic needs

Future dangers

Unaligned superintelligence

If you're not familiar with this, you can start by watching: If anyone builds it, everyone dies

Now is the time to figure out how to prevent catastrophic AI, before it's too late. If you have policy ideas or more, join the discussion.

@@ Line 1: / Line 1: @@
-==Current threats==
+==Current dangers==
-===Not being able to trust camera footage & audio recordings===
+{{R|Why it matters
-* Could make it harder to solve crimes
+|1. Don't let governments become dictatorships. Footage is a key part of holding authorities accountable. Who will watch the watchers? We the people. We can't let AI ruin that.
-** This includes [[police brutality]]. Police & governments must be held accountable. The fact that everyone has phones and dashcams has helped a lot this past decade. Now this might be at risk.
+}}
-* Voice impersonation scams
+===Not being able to trust pictures, video, audio===
-* Misleading crap on the internet {{light|(political disinformation, etc)}}
+{{R|FAQ
-====What could be done about it? {{fists}}====
+|''What about AI detection software?''<br />Even though generators and detectors both get more advanced over time, the generators ultimately win, unfortunately. ''Read&nbsp;more''
-* Content signatures?
+}}
-* Mandatory watermarks?
+What could be done:
-* Enact some new policies before it's too late?
+:; policy ideas:
-* ''Got ideas? Join the '''{{tp2}}'''''
+:* [[Ban unwatermarked generative AI]] {{talk|should watermarking be combined with an additional policy where every AI image generator company has to keep a database of all generated images, and offer a free service for anyone to verify any image on the internet, as in, to see if it matches something in the database?<br />what should we title such a policy?}}
+:; tech ideas:
+:* [[Anonymous camera digital signature]]s
-===Chatbots wreaking havoc on mental health===
+Status quo:<small>
-{{empty}}
+* Google's AI image generators produce images with a "synth ID" so they can be identified as fake. Not sure how foolproof it is though.
+* Sora AI (video generator) adds a subtle watermark to the video. Some people have been known to crop it out though.
-<!-- MAYBE:
+* There are plenty of image generators that don't produce any watermark at all.
-===Spam and slop===
+* Watermarks in general are not hard to remove, unfortunately. Still better than nothing though; most people probably won't bother removing them. So watermarking can at least reduce the total amount of misleading slop on the internet.</small>
-* Scams can be made to look more legit.
+{{R|FAQ
-* Far-right political bots trolling the internet, wasting the time of people who actually care about discussing how to build a better world.
+|Why not just create more jobs? {{p2|''Read&nbsp;more''
-* Hard to promote good written content in a sea of generated crap. People might think your writing is AI anyway
+|History is full of technological breakthroughs that made work more efficient. But instead of people working less, the economy adapted<!--responded--> by ''producing more.'' That's what they call "economic growth", and it's been one of the most ''inefficient'' ways to lift people out of poverty. After all these decades, most people are still overworked at their jobs, and there are still almost a billion people malnourished. We need to do better.}} }}
--->
+{{R|Other concerns
+|* [[Environmental impact of AI]]<br />* [[AI psychosis]]
-===What about unemployment?===
+}}
-[[File:Robots under capitalism vs socialism.jpg|thumb]]
+{{R|Other discussions
-Unemployment is only a problem if we stick with the old capitalist system of needing a job to afford food, housing, etc.
+|* [[Talk: Are AI images art or plagiarism?]]
-* Time to [[socialize basic needs]]<!--ALT: '''[[Universal basic needs|There are better alternatives]]''' to the current system.-->
+}}
-* <small>Also note: [[We could be working less, even without AI]]</small>
+===Unemployment===
+Main solution: [[Socialize basic needs]]
-==Future threats==
+==Future dangers==
 ===Unaligned superintelligence===
-{{empty}}
+: <small>''If you're not familiar with this, you can start by watching: '''[[If anyone builds it, everyone dies]]'''''</small>
+Now is the time to figure out how to prevent catastrophic AI, before it's too late. If you have policy ideas or more, join the {{tp2}}.
-Help build this section - join the [[{{TALKPAGENAME}}|'''discussion''']].
-==See also==
-* '''[[Environmental impact of AI]]'''
-* [[Talk: Are AI images art or plagiarism?]]