There’s a tendency in this heated political climate to simply reject people who are saying false things and to write off conspiracy theorists writ large.
But as the US approaches the third straight election in which misinformation — and the fight against it — is expected to play a role, it’s important to understand what’s driving people who don’t believe in US elections.
…
I talked to O’Sullivan about the documentary, in which he has some frank and disarming talks with people about what has shaken their belief in the US. But he paints an alarming picture about the rise of fringe movements in the country.
Our conversation, conducted by phone and edited for length, is below:
WOLF: What were you trying to accomplish with this project?
O’SULLIVAN: So much of mainstream American politics now is being infected and affected by what is happening on what was once considered the real fringes — fringe platforms, fringe personalities.
And I think really what we want to do in this show is illustrate how these personalities may be pushing falsehoods, but they’re no longer fringe. This is all happening right now. And it is having a big effect on our democracy.
…
I was about to say the same thing applies to AI. But AI is fucked right out of the gate. There’s not even a brief window of hope for it being used to better society. Anyone with any awareness on the topic knows these AIs are already corrupted and compromised because they’ve been using the Internet to train all their LLMs.
Have you actually used them? I’ve been shocked at how well they respond to almost everything I’ve thrown at them. From my experience, they give accurate medical advice, they demonstrate appropriate emotional intelligence, they are good at coding, they give sound advice, they’re good at summarizing, they can write complicated papers on niche topics, the list goes on. I’ve never encountered something controversial or offensive, nor anything bigoted, or racist, or ignorant, or elitist, etc. LLMs are shockingly robust and, quite frankly, incredible at what they do.
Yes, I’ve used them. And for some things they are okay (for example, their responses on programming questions can be hit or miss).
But given that there have been active disinformation campaigns and election meddling attempts (with several countries being both the victims and the perpetrators of this), I’m not convinced they will be reliable on some topics. Not least because the companies behind them have been slow to take the topic of information fidelity seriously.
Well if you go back to the use of algorithms they did have this massive potential but they all to quickly got involved with advertising and social media and yeah. it was yuck already at that point. But like computer vision and such gave it so much promise.
algorithms could help people. they could, for example, help you find relaly cool obscure stuff on netflix/spotify that you might like. They worked like this for awhile and ti was great!
But that doesn’t make money. the algorithm that shoves netflix’s latest trash content does, so that is why it shows up in every suggestion an takes up so my screen space. the vast majority of my spotify ‘feed’ is podcast trash i have never or never ever will listen to, and i can no longer use it to find some obscure band playing weird music like I did 8 years ago.