I was curious what the context was. I went to the recording and transcribed:
“What’s a picture of where you’re hiring from, what are their backgrounds, kind of what brings them—”
“—College grads who go to generically, like, what you’d consider top 25 schools, who are interested in effective altruism. The same is basically true, mostly true, of the operations staff, the way that we’ve hired them, primarily entry level generalists who’ve come on and helped out. In both cases, we, GiveWell, know more about what types of people we need and what roles we need to fill. Our hope going forward, our plan, is not not hire as many of that type of person, and instead hire someone to fill specific roles. So on the operations side, instead of hiring a generalist to handle accounting and the accounting side of what we do, hire someone who has that kind of accounting background, because now it is a full-time role, in a way it wasn’t a couple years ago. On the research side a lot of what we do is assess academic literature, and we think we can find someone who knows that stuff cold because of their background, like being a PhD economist. That’s the direction I think we’re heading with the GiveWell—”
I liked Issa’s notes, but this reduces my confidence that they’re a good idea in their current form. Heavily paraphrasing unscripted off-the-cuff remarks in a mildly hard-to-parse, very condensed / nuance-stripped form seems like a recipe for starting misunderstandings and rumors.
I do agree that partial sentences can be harder to parse in some cases. However, in this case, I think Issa’s condensation of the transcript did not lose the spirit. I can see how some people would have taken more time to parse his shorter version, but I think it wasn’t more prone to misinterpretation than the full transcript (in other words, it didn’t add to the issue of misinterpretation). In particular, I don’t see any nuance in the original transcript that was missing from Issa’s condensation.
ETA: As disclosed at the end of the post, I sponsored its writing and provided feedback. To the extent that I didn’t ask Issa to expand that section of the transcript, it shows that, even prior to publication, I thought it was reasonably clear.
I think the transcript and summary will read the same to a lot of people, and read differently to a lot of other people. Connotation and wording is a complicated thing, and a lot of nuance and tone is already lost just in going from spoken conversation to written transcript, even before any summarizing or paraphrasing occurs. I’m not super concerned about this specific incident, but it brought to my attention how likely this is to be a problem going forward.
I was curious what the context was. I went to the recording and transcribed:
“What’s a picture of where you’re hiring from, what are their backgrounds, kind of what brings them—”
“—College grads who go to generically, like, what you’d consider top 25 schools, who are interested in effective altruism. The same is basically true, mostly true, of the operations staff, the way that we’ve hired them, primarily entry level generalists who’ve come on and helped out. In both cases, we, GiveWell, know more about what types of people we need and what roles we need to fill. Our hope going forward, our plan, is not not hire as many of that type of person, and instead hire someone to fill specific roles. So on the operations side, instead of hiring a generalist to handle accounting and the accounting side of what we do, hire someone who has that kind of accounting background, because now it is a full-time role, in a way it wasn’t a couple years ago. On the research side a lot of what we do is assess academic literature, and we think we can find someone who knows that stuff cold because of their background, like being a PhD economist. That’s the direction I think we’re heading with the GiveWell—”
I liked Issa’s notes, but this reduces my confidence that they’re a good idea in their current form. Heavily paraphrasing unscripted off-the-cuff remarks in a mildly hard-to-parse, very condensed / nuance-stripped form seems like a recipe for starting misunderstandings and rumors.
I do agree that partial sentences can be harder to parse in some cases. However, in this case, I think Issa’s condensation of the transcript did not lose the spirit. I can see how some people would have taken more time to parse his shorter version, but I think it wasn’t more prone to misinterpretation than the full transcript (in other words, it didn’t add to the issue of misinterpretation). In particular, I don’t see any nuance in the original transcript that was missing from Issa’s condensation.
ETA: As disclosed at the end of the post, I sponsored its writing and provided feedback. To the extent that I didn’t ask Issa to expand that section of the transcript, it shows that, even prior to publication, I thought it was reasonably clear.
I think the transcript and summary will read the same to a lot of people, and read differently to a lot of other people. Connotation and wording is a complicated thing, and a lot of nuance and tone is already lost just in going from spoken conversation to written transcript, even before any summarizing or paraphrasing occurs. I’m not super concerned about this specific incident, but it brought to my attention how likely this is to be a problem going forward.