This Week in Misinformation: The Prodigal Shaman, AI Unfixable (?), Fake AOC, Succession, Chick-Fil-A
1 June 2023
Keeping up on misinformation is basically the best thing you can do for your brain. So glad you’re here!
Give this newsletter a share, won’t you? Posting a snippet to social can be a good taste for people in your network, or sending this email directly also tends to work well. 🙂
Reliability scores for media outlets cited in the summary are in parentheses for each, courtesy of the terrific folks at Ad Fontes Media.
Now, on to our top stories.
The Shaman is back.
Jacob Chansley went to prison and then an Arizona halfway house after pleading guilty to crimes committed on January 6th, but now he’s served his time and is selling yoga pants with his costumed visage on them—and, generally, leaning into the impression he left as the horned, spear-wielding extremist Americans saw on their screens (VICE, 39.10).
His first public act as a free man was to speak at an event in Scottsdale (AZ Central, 43.97). The media initially reported he had appeared at a named church that had nothing to do with it (Fronteras, -), but in the AZ Central story above there is a correction of that mistake.
If the Shaman is trying to distance himself from any part of his previous life, it might actually be that whole “QAnon“ thing. Maybe he’s realized it was a massive LARP, about which we recently learned that the FBI decided it wasn’t worth bothering to look into (Insider, 42.04)?
Fixing AI’s misinformation problem is going to be difficult, bordering on impossible.
By now we all know that OpenAI’s large language model-powered ChatGPT “hallucinates”—read: makes things up—but have you heard that some researchers think that problem might never be fixable (Washington Post, 38.22)? Over at Google, an AI leader resigned amid splashy warnings about the dangers of AI, and hallucinations is one of the things he says should worry us (Fortune, 44.49).
We haven’t exhausted all ideas yet, though. For example, research suggests “bot debates” could produce better answers to prompts (Fortune). OpenAI announced it is pursuing a new way to combat hallucinations (CNBC, 44.82), so that’s something we’ll also be watching. Further: chatbots are not great at fact-checking (Poynter, 41.01), and pressure is rising on OpenAI to make ChatGPT's limitations clearer (The Verge, 44.19).
Visual synthetic media, meanwhile, is beginning to be used by major national political campaigns (Forbes, 41.08), and deepfakes represent a growing threat to the 2024 election (Reuters, 47.27). It’s a veritable tsunami crashing over us.
Elon’s Twitter censors, but who is doing the censoring is shifting.
Responding to what I’m not sure, Musk tweeted that Twitter has no “actual choice” but to censor what government officials request be censored (Insider), and to be fair the EU didn't take kindly to his decision to not participate in a related voluntary framework (Politico, 42.63). Elon did pledge to publish “all government censorship requests” going forward (@TitterDaily via Twitter), noting that all the social media companies do governments’ bidding.
Trust and safety head Ella Irwin resigned a bit abruptly (Reuters) but that possibly had more to do with today’s spat between right-wing blue checks and Twitter over the Twitter premier of a Daily Wire (32.95) film than anything Musk says or thinks about the limits of free speech.
Alternate theory: Irwin got caught up in the saga of the fake AOC account that Twitter helped amplify to the point it was seen more than the Congresswoman’s authentic one (NBC News, 45.02).
The ever-loving grab bag to finish us out: meteorologists are now the targets of conspiracy theorists; Succession fans should know this old baseball player was not a spoiler; chicken sandwich restaurant chain Chick-Fil-A did not recently create a DEI position, but that isn’t stopping conservative boycotts; the supervisor in Arizona’s largest county won’t run for re-election because of all the harassment; count yourself lucky if you’ve never heard of Trump Bucks; the American Greatness (18.87) reporter who will now have access to the full trove of January 6th footage announced her intention to "control the narrative" with it; Ron DeSantis says it’s weird that Donald Trump didn't destroy the Deep State despite having four years in the White House to do so; there was a cool Nobel Prize summit on the problem of misinformation last week (your editor was there!); and the American Hospital Association announces a strategy to tackle public health misinformation.
All that, and a lot more, below. This is This Week in Misinformation.
-- Kevin