3 ms·
Somewhat agree with your take. A good example is Twitter: it’s trivial even for a non-state actor to automate a bot farm to post hundreds of thousands of disinf
by fanso99 4y ago
Somewhat agree with your take. A good example is Twitter: it’s trivial even for a non-state actor to automate a bot farm to post hundreds of thousands of disinfo tweets, create “connections”, retweet each other, etc. this is happening today at a massive scale.
However, where I can see a model like ChatGPT making a difference is creating a tsunami of seemingly in-depth content both in public (Wikipedia) and private (buying 10K domain names) spaces. Generating fake research papers, interviews, blog posts, reviews… at a scale that until now required an unattainable amount of human supervision.
How much content would it take to flood the more visible part of the internet?
- cwkoss 4y agoFair point. But hasn't SEO blogspam already flooded every inch of the visible part of the internet already? Can't partisans already get 5-10 pundits to write articles advocating a questionable concept to add a veneer of credibility for a couple thousand bucks? I'm not sure what frontiers of the visible part of the internet remain. It will certainly become more economical to do this sort of nefarious stuff, but I suspect democratizing misinformation production will lead to the development of better media literacy at a rate that outpaces that misinformation drowning out people who are earnests trying their best to find and present the truth.
- fanso99 4y agoIn my experience blogspam has been primarily used for commercial purposes (selling stuff and/or ads). Political disinfo at a new scale could pose very interesting new challenges. In a similar way we saw a paradigm shift when Cambridge Analytica employed political targeting at a new scale. Before that anyone could hire a savvy social media manager to target specific groups, but automating this process proved to be a different beast. How soon our society will adapt to this, or how we will even react is yet to be seen. I don’t have high hopes for this to be a fast and painless process.
- commandlinefan 4y ago> to post hundreds of thousands of disinfo tweets If that's true, then they've been doing it for a long time now... has it actually made a difference? There have been a lot of news items that were accused of being disinfo that have since turned out to be absolutely true, but I have yet to have seen an example of actual disinfo that took hold and altered any meaningful course of events.