Solving Problems of the Pause AI movement

Hello. I want to help PauseAI. So i want you guys to describe the biggest problems in your efforts. The biggest problems, needs. Biggest problem, that happen most often, repeatedly.

I want to know those problems, in order to help solve them. Be it helping make strategies, or coding up tools, software, etc.

So please, write about problems you meet, when you are advancing the PauseAI goals, priorities. The more you guys write, the better, would help me better.

ill-bronzeOP•12/2/23, 9:35 PM

For example:
I find it hard to convince other people of risks of AI....

I find it hard to make other people here do helpful things....

I find it hard to find a common ground in the arguments...

I find it hard to see what actions would be helpful...

etc.

Iill-bronze For example: I find it hard to convince other people of risks of AI.... I find ...

popular-magenta•12/2/23, 10:29 PM

Same. People are complacent, ignorant (I mean that in the nicest way possible), assume AI risks are sci fi or comical, etc. Convincing people that it is actually able to harm us, that it's soon, and that it's likely are all difficult for me.

popular-magenta•12/2/23, 10:30 PM

They can't really envision how GPT can hurt them.

ill-bronzeOP•12/2/23, 10:55 PM

I see. Can you expand on that?

popular-magenta•12/2/23, 10:55 PM

They're convinced it's 100 years away or that I'm afraid of Arnold from The Terminator coming to kill me

popular-magenta•12/2/23, 10:56 PM

Additionally, just telling them it will be really smart doesn't translate in their head to dangerous. They just think we can turn it back off whenever and that will always be the case.

popular-magenta•12/2/23, 10:58 PM

I guess I don't know how to be more expansive without just listing examples of this kind of dismissive thinking about it, or normalcy bias

hurt-tomato•12/5/23, 1:32 PM

1. It's hard to get volunteers to act. With act, I mean things like sending emails, protesting, doing presentations, writing an article, etc. This is the most pressing problem from my perspective, because it should accelerate all other problems.
1. Feeling powerless. With climate change, we can turn down the heating, fly less often, vote for a green party, etc... We have all these small things that give us some way of helping out. With AI risk, this seems harder to convey. People don't know what to do, they don't know how they can help. "I'm not an expert / lobbyist / ML engineer / policy writer / debater". This is such a shame, because I think there are so many things that people can do! Part of why I do these email workshops, for example. The problem is almost NEVER capabilities - it's almost always a feeling of powerlessness or a lack of "just fucking do it" mentality.
1. Lack of ownership. People don't feel like they have responsibilty in all of this. AI risk is nobodies problem, until it's everyone's problem. I personally feel a lot of responsibility, which is why I started all of this. But I want other people to feel this. I'm trying to more explicitly give people tasks (e.g. new teams / working groups), and this does help, but it's nowhere near as much as I'd like.
1. The psychology. X-risk is hard to talk about, hard to understand, hard to believe and hard to act on. Read more: https://pauseai.info/psychology-of-x-risk. I find this more important than the technical argument.

Hhurt-tomato 1. **It's hard to get volunteers to act**. With act, I mean things like sending ...

ill-bronzeOP•12/6/23, 4:08 AM

Thanks for writing this

opposite-copper•3/17/24, 7:31 AM

It sits at a weird balance of uncertainty for me.

opposite-copper•3/17/24, 7:32 AM

Like, I know the problems are real, but I get so uncertain what to do about it that I can go for long stretches not obviously doing anything at all.

opposite-copper•3/17/24, 7:33 AM

And I want to "do a really good job on the AI safety stuff" so I procrastinate it and get caught up in stuff I feel like I am actually allowed to do instead.

opposite-copper•3/17/24, 7:36 AM

For me, I think my next concrete target with PauseAI is to contact (call or email?) my representatives (found with https://whoaremyrepresentatives.org/) and tell them that I am concerned about the existential and humanitarian risks of AI and that I think there should be a Pause and a global treaty/summit of some kind.

opposite-copper•3/17/24, 7:38 AM

It is intimidating to me because I am not politically trained and have no idea how to "lobby" lol.

Oopposite-copper It is intimidating to me because I am not politically trained and have no idea h...

hurt-tomato•3/17/24, 9:37 AM

You can do this! You don't have to be a professional lobbyist - politicians love to talk with people who are not professionals. Last week I met one who said what a relief it was to speak to someone who's NOT from big tech!

You just talk about the problems, help them understand what needs to be done.

hurt-tomato•3/17/24, 9:38 AM

However, keep in mind that they are just people too. Establish a connection. Ask about their day, about their work.

hurt-tomato•3/17/24, 9:38 AM

Don't barge in and give a lecture

hurt-tomato•3/17/24, 9:38 AM

Find a connection

dead-brown•3/18/24, 9:44 AM

!track

ProjectBot•3/18/24, 9:44 AM

Project Tracker: https://maximolog.notion.site/Solving-Problems-of-the-Pause-AI-movement-efa022458c064f51b8fbee9f01e61811

Maxime Fournes's Notion on Notion

Solving Problems of the Pause AI movement | Notion

Solving Problems of the Pause AI movement

Similar Threads

Similar Threads

Similar Threads