July 2024 Social Media Posts

Help with writing captions for posts that will draw new people into the PauseAI movement and support our social media presence as an entity.

popular-magentaOP•6/24/24, 6:39 PM

I have a bunch of social media posts almost fully designed, but coming up with clever, thoughtful, attached captions is just taking too much time away from other initiatives. For this project, we're trying to write one or two sentences for each of about 30 of these posts calling out risks of AI that the public might not be familiar with.

popular-magentaOP•6/24/24, 6:41 PM

For each post, reply in this channel with the number of the post, followed by the caption you've thought of

popular-magentaOP•6/24/24, 6:41 PM

1-9

popular-magentaOP•6/24/24, 6:42 PM

10-18

popular-magentaOP•6/24/24, 6:42 PM

19-26

popular-magentaOP•6/24/24, 6:42 PM

27-29

popular-magentaOP•6/24/24, 6:46 PM

Also, if you want changes to the content, let me know

standard-azure•6/24/24, 9:30 PM

Writing so that you know it wasn't pointless to set this project - I want to contribute, but right now I am to tired. I hope not forget to write after sleep.

awake-maroon•6/24/24, 11:13 PM

I think it makes sense to at least focus on issues caused by ai being good, not by it being bad. like "errored on 90 % of claims": cool, so let’s make it better

Aawake-maroon I think it makes sense to at least focus on issues caused by ai being good, not ...

popular-magentaOP•6/25/24, 1:55 AM

Okay, that's a fair point. I guess my point with that one (and a few others) is to show that companies will turn AI superhuman AI against people, using the biases and failures as a feature to make them more money while they still claim to be 'fair'. The word 'errored' should probably be replaced with something else.

radical-lime•6/25/24, 5:38 PM

I just want to say these designs are SO GOOD. They're eye-catching, informative, and straight to the point. These are exactly the kind of posts that can get a lot of visibility and engagement.
That being said, why stop at just 1-2 sentences in the captions? I think a lot of these could benefit from a longer caption that elaborates on the issue, maybe even with some stats. I've seen this format on a lot of similar posts and it seems to work. Would help with spreading good information but also lower the chances of people thinking it's just fearmongering or something

awake-maroon•6/25/24, 6:45 PM

Hm, idk if rewording fixes the issue

Rradical-lime I just want to say these designs are SO GOOD. They're eye-catching, informative,...

popular-magentaOP•6/25/24, 7:05 PM

That's what this project is - writing text captions

Aawake-maroon Hm, idk if rewording fixes the issue

popular-magentaOP•6/25/24, 7:06 PM

Yes, refocus on the number of claims denied or something. Or I'll pick a different issue. We don't have to do that one.

radical-lime•6/25/24, 9:51 PM

Oh of course, I just thought you wanted to do 1-2 sentences only for the captions. Maybe I misunderstood

Rradical-lime Oh of course, I just thought you wanted to do 1-2 sentences only for the caption...

popular-magentaOP•6/26/24, 2:06 AM

Oh, my bad. I mean, I was only planning to do 1-2 sentence captions to keep them condensed and not spend a million hours of people's time on them. But if you want to write longer ones and think they're better, that's great.

radical-lime•6/26/24, 3:02 AM

I haven't been approved as a volunteer yet (interview in a couple days) but I would love to help if I am!

Rradical-lime I haven't been approved as a volunteer yet (interview in a couple days) but I wo...

popular-magentaOP•6/26/24, 3:08 AM

You don't need to be approved to do this

popular-magentaOP•6/26/24, 3:09 AM

And I'm not sure who you're interviewing, because I never interviewed anyone

radical-lime•6/26/24, 3:10 AM

Oh? I filled out the volunteer form and I have a meeting booked, I guess that was my assumption

radical-lime•6/26/24, 3:11 AM

Maybe it's just to help me figure out where I can help most

radical-lime•6/26/24, 3:13 AM

Well in that case I can just start drafting a bit at the very least :) did you have a deadline in mind? Also this is for Instagram, right?

Rradical-lime Well in that case I can just start drafting a bit at the very least :) did you h...

popular-magentaOP•6/26/24, 3:20 AM

Yeah, probably. But this works too.

popular-magentaOP•6/26/24, 3:20 AM

I was hoping to post some of these in July. So if we get at least 10 fully done by then, that will be great.

radical-lime•6/26/24, 4:08 AM

Gotcha! I'll start working on some

popular-magentaOP•6/26/24, 12:52 PM

Great, thanks

radical-lime•6/28/24, 1:40 AM

Hey! So I've been drafting a few here and there but I just finished #6, let me know what you think. I know it may seem too long but bear with me

radical-lime•6/28/24, 1:43 AM

3 seconds. "Hello?" "Yes." "Who's calling?" That's all it takes. Identity theft, fraud, and impersonation scams are nothing new, but AI's rapid advancements and ease of accessibility to even the least technical users has fielded an uptick in the rate of these crimes at an alarming rate.

As a result, 73% of Americans are concerned about AI-generated deepfake robocalls mimicking the voice of a loved one to scam them.

Here are 5 ways to protect YOUR voice from being used in a cloning scam. Share this with your friends and family!

SWITCH to automated voicemail messages. Personalized greetings are long enough for bad actors to capture.
CREATE a family safe word. Impersonating a family member is the most common cloning scam. Consider agreeing on a safe word with your loved ones to use in the case of emergencies or suspected cloning.
LIMIT social media recordings. Be mindful about what you post, especially if you share your voice on public videos. Lengthy and various recordings of your voice make it easier for bad actors to clone it. Phrases like "help me," "I'm hurt," "stop," etc. can be very easy to manipulate to make it sound like you're in danger.
AVOID voice biometric verification. When voice samples are saved during the verification process, they can be easy targets for bad actors.
DO NOT speak first to unknown callers! Remember, bad actors only need a few seconds of your voice. Words and phrases to avoid include "yes" if an unknown caller asks you if they're speaking to you.

Don't fall victim to these scammers. Protect yourself and your loved ones by making these small changes today.

popular-magentaOP•6/28/24, 11:45 AM

Okay, that was longer than I was expecting

popular-magentaOP•6/28/24, 11:46 AM

Don't go much longer than this, because there's a 2200-character limit on Instagram

popular-magentaOP•6/28/24, 11:48 AM

Also, we need to make sure we keep coming back to Pause. Personal protection is important, but that's not really our priority.

radical-lime•6/28/24, 5:42 PM

Sure, but people should be equipped to do what they need to do to protect themselves esp if the post references something like that. Otherwise, we could get a lot of comments from people asking what they can do or people being turned away because we aren't actually providing solutions. I see that problem a lot on activist accounts that aren't doing too well. I also used to run one.

radical-lime•6/28/24, 5:44 PM

That being said the length can be trimmed down in a lot of places, i.e. the "why" in the steps to take aren't completely necessary. That leaves room for something about Pause, but maybe you can give me an idea of what you want regarding that?

Rradical-lime Sure, but people should be equipped to do what they need to do to protect themse...

popular-magentaOP•6/28/24, 8:38 PM

If you've seen the issue on activist accounts before, that's enough justification for me. It's more that I've been retooling many of my projects toward more existential risks at the request of other members, and I was preventatively working on this one. But if there's a reason not to, that's fine.

popular-magentaOP•6/28/24, 8:39 PM

Though I still do think something about 'This will get worse with stronger AI models, join PauseAI to protect thousands of people from nightmarish spam calls." would be good to put at the end.

popular-magentaOP•6/28/24, 8:40 PM

And do you think you can do things of part of this length for all of them? If this one is a bit longer that's fine, but I can't imagine wild swings in length are good for consistency,

radical-lime•6/28/24, 10:46 PM

That's perfect, I'll add that at the end and trim the rest down
Also I'll definitely do my best to keep everything the same length :)

radical-lime•6/28/24, 10:51 PM

Maybe something like this instead?

3 seconds. That's all it takes. Identity theft, fraud, and impersonation scams are nothing new, but AI's rapid advancements and ease of accessibility to even the least technical users has fielded an uptick in the rate of these crimes at an alarming rate.

SWITCH to automated voicemail messages.
CREATE a family safe word in the case of an emergency.
LIMIT social media recordings featuring your voice.
AVOID voice biometric verification.
DO NOT speak first to unknown callers!

This will lower the risk FOR NOW! As AI models get stronger, this will get worse and harder to combat.
Join PauseAI today to protect yourself, your loved ones, and thousands of people from nightmarish spam calls.

popular-magentaOP•6/28/24, 11:02 PM

Okay, I'll adjust this and attach it.

radical-lime•6/29/24, 12:52 AM

Lmk if you post it :) I'll keep working on some other ones

popular-magentaOP•6/29/24, 3:35 AM

I'll post it eventually, not until we have a few more so they can be somewhat consistently timed.

popular-magentaOP•6/29/24, 3:36 AM

But also, anyone can post them. We're not holding them back.

radical-lime•6/29/24, 6:17 AM

Is there an official account or is this something we can just post to our own accounts individually?

Rradical-lime Is there an official account or is this something we can just post to our own ac...

popular-magentaOP•6/29/24, 12:01 PM

There are a few PauseAI and PauseAI regional accounts that we use.

popular-magentaOP•7/3/24, 5:28 PM

@Malyssa Any update on this?

July 2024 Social Media Posts

Similar Threads

Similar Threads

Similar Threads