Safety experts quit OpenAI
- Thread starter Agema
- Start date
You can say like the same thing about people's estimates of anything and this is without AI. Maybe even more of a stream of information may tilt the numbers more but not much, overestimating too much just doesn't make sense. You think with AI pummeling the British with more immigration stories is going to result in people thinking over half the population is immigrants? That likely makes no sense.
Although, 41% of democrats thought an unvaccinated person had a 50% or higher chance of being hospitalized if they got covid, that's pretty ridiculous. And, again, this is without AI so what's so dangerous about AI?
U.S. Adults' Estimates of COVID-19 Hospitalization Risk
Americans' estimates of both the risk of hospitalization due to COVID-19 and the efficacy of the vaccine for the disease vary widely by their political affiliation and vaccination status.
news.gallup.com
They're not AI, even by the broader definition. They're essentially search engines with an executive file attached to them. They have no truly adaptive, learning element to them.
What people think already doesn't 'make sense'. That's not a barrier. Yes, volume (as well as sophisticated targeting and quick responsiveness to trends) matters.
So let's say the misinfo on immigration hits an upper limit on how far peoples' credulity will go. Human-made misinformation took thousands of work hours, tens of years, to create this false impression. AI can churn out the same volume in days... on a hundred different topics. So people cease to believe immigration lies once they go over a certain scope? Then misinfo targets other questions altogether: 50 at once, and creates more than enough volume in a tiny fraction of the time.
Humans have simply been unable to create, curate, and target the volume of misinformation, on the same timetrame. That's unarguable.
Since I'm not in Britain, I'd be assuming the reason for the higher percentage is because the sheer amount of news stories about migration. And also humans are kinda bad with percentages, there's a jackbox game about that in fact. I doubt the percentage would go much higher with AI pushing out even more stories.
Reread what specifically? You can only flood people with so much info, the person is in essence the bottleneck (like a real hard drive with a slow spin rate and very fragmented). People need to see enough stories about something and also see the actual effects of it. You tell me there's a bear problem in my city and I'm not gonna believe it until I see a bear or someone I know has seen a bear or knows someone that's seen a bear. We have that process basically as sped up as possible already. People don't need all these man hours to spread this information, you have have one story, then other news publications basically copy/paste it and then you have have news anchors on TV/Youtube or influencers on TikTok or whatever, and bots spreading essentially that one main story.
For people to believe there's an immigration problem, it's not gonna happen in one day or a few hours because of AI, it'll still take the normal cycle for it to become a thing.
The words.
Bollocks. Constant, low-substance content can be consumed or absorbed nearly endlessly-- and if there's an upper limit, we're not there, even in today's bombardment. And people absolutely do not need to see the actual effects of something to have strong opinions on it-- hence why endless stories of fabricated nonsense like EU banana-straightening helped to swing a nationwide referendum. There's a reason that the people who tend to have the biggest problems with immigration tend to live in low-immigration areas.
Everything can be done with bots, you don't need AI. We already have like one or two sources that actually get the story and the rest regurgitate it. I don't know how you're getting thousands of hours of work out of this.
Can be done with bots or humans, with a lot more time and human. That's the point. Scale and speed is the point. And it's not comparable.
I'm not going to be consuming any more news content than I already am (I'm guessing the same is the case for most people). How is AI going to force me to consume all the more stories it can make?
When someone just endlessly scrolls through Reddit or Twitter or Instagram or whatever, how much content do you think they're properly consuming? And how much do you think they're passively aware of?
I don't doom scroll on any of those sites/apps. Also, just because people scroll Reddit or whatever doesn't mean they are scrolling news content.
I don't care what you're doing. Hundreds of millions of people do, including news content. How much do you think they're presented with, or passively take notice of, even if they don't actively open and read an article?
Then they are idiots. When I am bored I go on Tik Tok, when I am serious I go to more established news sources. Your BBCs, New York Times, and etc.
I'm sure people have a time where they check out news regardless of source (could be MSN, Reddit, Twitter, TikTok, Youtube, etc), then they have times to check general interests / hobby stuff / memes / etc. Why would you need to flood the internet with more news content? There's already plenty of it.
OK. You may think you're immune to misinformation, but you're not. And even of it doesn't have much effect on you... the elections in your country will be decided, in part, by people who have been effected.
Frankly, I think you just have an extremely simplistic view of how media is widely consumed.