I liked the article on a personal level, but as PR I agree more with Kawoomba. It seems like a lot of people invested time into making this article well-informed and balanced, yet the result is a (mild) PR net-negative, albeit an entertaining one. We have positive associations with most of the things the article talks about, so we’re likely to underestimate the effect of the article’s negative priming and framing on a typical reader (which may include other journalists, and thereby affect our perception in future articles).
The article wants readers to think that Vassar has delusions of grandeur and Thiel is a fascist, so linking the two more tightly isn’t necessarily an effort to make either one look better. And there’s totally such a thing as bad press, especially when your main goal is to sway computer science types, not get the general public to notice you. This at best adds noise and shiny distractions to attempts to talk about LW/MIRI’s AI views around the water cooler.
It’s true FHI and FLI are more PR-oriented than MIRI, but that doesn’t mean it’s FHI’s job to produce useful news stories and MIRI’s job to produce silly or harmful ones. Better to just not make headlines (and reduce the risk of inoculating people against MIRI’s substantive ideas), unless there’s a plausible causal pathway from the article to ‘AGI risk is reduced’.
We have positive associations with most of the things the article talks about, so we’re likely to underestimate the effect of the article’s negative priming and framing on a typical reader
I think hostile media bias is stronger. Priors indicate that the average person on LW while have a more negative view of the article than warranted.
The article wants readers to think that Vassar has delusions of grandeur
I don’t think the average person watching Vassar’s Tedx talk would get the same impression as someone reading that article.
Thiel is a fascist
If that’s what he wanted to do he would have made a point about what Palantir Technologies does. Maybe remind the readers about Palantir Technologies responsiblity for the attempt to smear Glenn Greenwald and destroy his career.
Instead the author just points out that Palantir is a nerdy name that comes from Lord of the Rings. Even if the author hasn’t heard of the episode with Glenn Greenwald, leaving out that Palantir is a defensive contractor that builds software for the NSA is a conscious choice that someone who wanted to portray Thiel as a facist wouldn’t make. The author went for “bunch of strange nerds” instead of “facists”.
And there’s totally such a thing as bad press, especially when your main goal is to sway computer science types
I don’t think that’s the audience that Harper’s magazine has. That’s not for whom a journalist in that magazine writes.
Did you had any negative water cooler discussions with people because of that article?
I liked the article on a personal level, but as PR I agree more with Kawoomba. It seems like a lot of people invested time into making this article well-informed and balanced, yet the result is a (mild) PR net-negative, albeit an entertaining one. We have positive associations with most of the things the article talks about, so we’re likely to underestimate the effect of the article’s negative priming and framing on a typical reader (which may include other journalists, and thereby affect our perception in future articles).
The article wants readers to think that Vassar has delusions of grandeur and Thiel is a fascist, so linking the two more tightly isn’t necessarily an effort to make either one look better. And there’s totally such a thing as bad press, especially when your main goal is to sway computer science types, not get the general public to notice you. This at best adds noise and shiny distractions to attempts to talk about LW/MIRI’s AI views around the water cooler.
It’s true FHI and FLI are more PR-oriented than MIRI, but that doesn’t mean it’s FHI’s job to produce useful news stories and MIRI’s job to produce silly or harmful ones. Better to just not make headlines (and reduce the risk of inoculating people against MIRI’s substantive ideas), unless there’s a plausible causal pathway from the article to ‘AGI risk is reduced’.
I think hostile media bias is stronger. Priors indicate that the average person on LW while have a more negative view of the article than warranted.
I don’t think the average person watching Vassar’s Tedx talk would get the same impression as someone reading that article.
If that’s what he wanted to do he would have made a point about what Palantir Technologies does. Maybe remind the readers about Palantir Technologies responsiblity for the attempt to smear Glenn Greenwald and destroy his career.
Instead the author just points out that Palantir is a nerdy name that comes from Lord of the Rings. Even if the author hasn’t heard of the episode with Glenn Greenwald, leaving out that Palantir is a defensive contractor that builds software for the NSA is a conscious choice that someone who wanted to portray Thiel as a facist wouldn’t make. The author went for “bunch of strange nerds” instead of “facists”.
I don’t think that’s the audience that Harper’s magazine has. That’s not for whom a journalist in that magazine writes.
Did you had any negative water cooler discussions with people because of that article?