This week's dispatches from the Ministry of Intrigue
Hello, faithful reader.
We published the following fresh dispatches this week:
Quote: Simon Willison on AI Deepfake Meeting Attendees
June 6, 2024, 2:52 p.m.
Ars Technica includes a great quote from Simon Willison on Zoom CEO Eric Yuan’s vision of AI doubles attending meetings on your behalf.
“I’m not a fan of that idea where people build LLM systems that attempt to simulate individuals,” wrote AI researcher Simon Willison recently on X, independently of the news from Yuan. “The idea that an LLM can usefully predict a response from an individual seems so obviously wrong to me. It’s equivalent to getting business advice from a talented impersonator/improv artist: Just because they can ‘sound like’ someone doesn’t mean they can provide genuinely useful insight.”
— Benj Edwards, Zoom CEO envisions AI deepfakes attending meetings in your place
Quote: Meredith Whittake on AI
June 6, 2024, 2:42 p.m.
During a talk on the privacy implications of the AI data grab, Signal’s Meredith Whittake provides a quote for the ages on the current state of the technology.
“It costs hundreds of millions of dollars to train these models,” Whittaker said. “So there is deep pressure from companies — that are basically promising God and delivering email prompts — to make some return on investment in this technology.”
— Meredith Whittake, Signal’s Meredith Whittaker: AI is a privacy nightmare
You should read the whole article where for the numerous issues, but this gem needed special recognition.
Quote: 3M Hid Toxicity of Forever Chemicals
June 6, 2024, 2:10 p.m.
A bombshell report from the New Yorker on how 3M discovered the dangers of toxic forever chemicals, but then concealed this revelation while simultaneously increasing their use in products.
What Hansen didn’t know was that 3M had already conducted animal studies—two decades earlier. They had shown PFOS to be toxic, yet the results remained secret, even to many at the company. In one early experiment, conducted in the late seventies, a group of 3M scientists fed PFOS to rats on a daily basis. Starting at the second-lowest dose that the scientists tested, about ten milligrams for every kilogram of body weight, the rats showed signs of possible harm to their livers, and half of them died. At higher doses, every rat died. Soon afterward, 3M scientists found that a relatively low daily dose, 4.5 milligrams for every kilogram of body weight, could kill a monkey within weeks. (Based on this result, the chemical would currently fall into the highest of five toxicity levels recognized by the United Nations.) This daily dose of PFOS was orders of magnitude greater than the amount that the average person would ingest, but it was still relatively low—roughly comparable to the dose of aspirin in a standard tablet.
In 1979, an internal company report deemed PFOS “certainly more toxic than anticipated” and recommended longer-term studies. That year, 3M executives flew to San Francisco to consult Harold Hodge, a respected toxicologist. They told Hodge only part of what they knew: that PFOS had sickened and even killed laboratory animals, and had caused liver abnormalities in factory workers. According to a 3M document that was marked “confidential,” Hodge urged the executives to study whether the company’s fluorochemicals caused reproductive issues or cancer. After reviewing more data, he told one of them to find out whether the chemicals were present “in man,” and he added, “If the levels are high and widespread and the half-life is long, we could have a serious problem.” Yet Hodge’s warning was omitted from official meeting notes, and the company’s fluorochemical production increased over time.
— Sharon Lerner, The New Yorker How 3M Discovered, Then Concealed, the Dangers of Forever Chemicals
Another fine example of American capitalism posing existential risks to the species.
And that's it!
Grave dust and falling leaves.