This is a special post for quick takes by Bella. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
EAs are trying to win the "attention arms race" by not playing. I think this could be a mistake.
The founding ideas and culture of EA was created and “grew up” in the early 2010s, when online content consumption looked very different.
We’ve overall underreacted to shifts in the landscape of where people get ideas and how they engage with them.
As a result, we’ve fallen behind, and should consider making a push to bring our messaging and content delivery mechanisms in line with 2020s consumption.
Also, EA culture is dispositionally calm, rational, and dry.
This is a poor fit for getting any traction in the current attention landscape.
If we don’t adapt, we risk increasing irrelevance.
a) a lot of EA's core worldview philosophy is about doing boring stuff (that works), and so we attract people with an aesthetic repulsion to overmarketing that stick here and contribute highly. It's not clear that standard marketing strategies work for something like EA without making such people more likely to leave, so I would be hesitant to propose changes to the current setup.
b) the School of Moral Ambition is already doing essentially a more marketed version of EA. I highly recommend anyone interested in this hop over to their platforms to check it out.
c) Cause area marketing is going fairly well I think? You may wish to consider voting for One for the World in our donation election.
d) We're about to have a 3000-person conference this weekend, our largest ever. "Increasing irrelevance" EA is decidedly not. Clearly our current approach is doing something right.
I'll add onto c) that AI safety cause area marketing is going really well (to the point I'm personally uneasy about it), and animal advocacy cause area marketing also seems to be doing ok. It's not just GHD cause area marketing that's working.
My reservations about anti-marketing effects apply mostly to principles-first EA outreach.
The bad news is that saying just go and make high-quality video essays is like saying just go and make high-quality movies. Literally, some of these videos require a similar amount of work as making a microbudget indie movie. The main creator might work on them full-time for over a year, and often involve an editor and, in some cases, people to help with other aspects of production like music, visual effects, or voiceover. (ContraPoints lamented the fact that she spent longer writing the script for her video about Twilight than Stephanie Meyer spent writing Twilight.) The people who make these videos have honed their craft over years of experience.
Jenny Nicholson’s older videos, which are shorter and have less razzle dazzle, are much lighter on production, but still a lot of time and effort. Hank Green does 2-3 vlogs per week on his channel and the vlogbrothers channel that I like a lot, which are relatively quick to make (they have to be, to come out so frequently). But Hank Green has this amazing radio host, podcaster, YouTube vlogger, TV personality-type charisma that is rare and also not a thing where you can say just go do that. If you think you can go do that, then definitely do it! But it’s not something just anyone can do, and no one can do it easily.
Another positive example from YouTube: Kirk Honda, a clinical psychology professor whose channel is called Psychology in Seattle. He does unbelievably good educational videos on mental health, psychology, personality disorders, therapy, abusive relationships, and so on, and the spoonful of sugar that makes the medicine go down is he’s often reacting to reality TV. But this is not a gimmick. He says that for ethical reasons, clinical psychology professors don’t have recordings of real people interacting that they can show to their students. Without real (or realistic) examples, it’s hard to convey what you’re talking about. Reality TV is a great teaching tool because these Hollywood studios have decided to do the arguably unethical thing and exploit people’s anguish for money, so he can just comment on it and use it to explain psychology concepts. These videos seem relatively quick to produce, but again you have the Hank Green problem: Kirk Honda has charisma, wisdom, eloquence, and charm, and not just anyone can do it, and no one can do it easily.
Podcasts are another medium that is incredibly popular, but also hard to do. Effective altruism already has the 80,000 Hours Podcast which is by and large a great success. The winning formula for shows like that is just booking fascinating people with fascinating things to say and letting them talk for a long time. And having good audio and video production, and having an interviewer who can lightly steer the fascinating person into saying more fascinating things. This is the easiest formula to replicate, but 80,000 Hours is already doing it. Not that I discourage people from starting new podcasts, or that I think the 80,000 Hours Podcast can’t be improved — my suggestion to them was to shake things up with guests who say things other than what people in EA are used to hearing, e.g., Richard Sutton, Yann LeCun, Jeff Hawkins, Edan Meyer. Toby Ord’s episode of the 80,000 Hours podcast was an amazing example of this. Toby Ord is as central to EA as it gets, but the cold water he’s dumped on AI scaling very much goes against most people in EA are saying about AI.
I’ve worked on 3 podcasts as hobby projects, and let me tell you, unless you’re already an expert, audio production is surprisingly hard to work out. If you’ve got a budget of ~$200 per episode (take this with a grain of salt and double-check my recollection/estimate), you can take the vast majority of this complexity away by renting a podcast recording studio for 1-2 hours. I think there are probably multiple spaces you can rent in any large city in developed countries. (This requires the people on the podcast are all physically in the studio. I don’t have a solution for making it easier to record podcasts remotely.) Editing the audio isn’t nearly as complicated and just takes time and work, but if you’ve got more money to spend, you can also hire an editor to take that complexity off your hands as well.
The right way to think about this is you’re trying to make good art. Easier said than done. However, I think it’s worth thinking about, and it’s worth trying for people who are willing to put in what it takes to give it a good shot. I just want to discourage people from thinking about making “content” or “social media content” as opposed to art.
I think investing in [shortform] mediums would just be good money chasing after bad.
Interesting!! Curious for any more detail on why you think this, if it's not too annoying to write out :)
if you have in mind content like the highest-quality video essays on YouTube
Yep, that's one of the things I'd be super excited about!!
To that end, earlier this year I helped get started AI in Context, which has been heavily inspired by the awesome creators you mentioned above :)
It sure is time and resource-heavy to get the videos out (we've only managed two so far (working on the third!) even though we hired the first programme staff in early Feb), but my hope is that it's worth it — the reception has been broadly very positive :)
EAs are trying to win the "attention arms race" by not playing. I think this could be a mistake.
I see this take a lot.
My immediate response is fourfold:
a) a lot of EA's core worldview philosophy is about doing boring stuff (that works), and so we attract people with an aesthetic repulsion to overmarketing that stick here and contribute highly. It's not clear that standard marketing strategies work for something like EA without making such people more likely to leave, so I would be hesitant to propose changes to the current setup.
b) the School of Moral Ambition is already doing essentially a more marketed version of EA. I highly recommend anyone interested in this hop over to their platforms to check it out.
c) Cause area marketing is going fairly well I think? You may wish to consider voting for One for the World in our donation election.
d) We're about to have a 3000-person conference this weekend, our largest ever. "Increasing irrelevance" EA is decidedly not. Clearly our current approach is doing something right.
I'll add onto c) that AI safety cause area marketing is going really well (to the point I'm personally uneasy about it), and animal advocacy cause area marketing also seems to be doing ok. It's not just GHD cause area marketing that's working.
My reservations about anti-marketing effects apply mostly to principles-first EA outreach.
An additional reason EAs may not be playing the attention arms race is that they may be persuaded by the fidelity model of spreading ideas.
Whether this is true or not depends on what specifically you mean by it. If by attention-grabbing content you mean:
Then I think it’s not true. I think investing in those mediums would just be good money chasing after bad.
On the other hand, if you have in mind content like the highest-quality video essays on YouTube, such as:
The bad news is that saying just go and make high-quality video essays is like saying just go and make high-quality movies. Literally, some of these videos require a similar amount of work as making a microbudget indie movie. The main creator might work on them full-time for over a year, and often involve an editor and, in some cases, people to help with other aspects of production like music, visual effects, or voiceover. (ContraPoints lamented the fact that she spent longer writing the script for her video about Twilight than Stephanie Meyer spent writing Twilight.) The people who make these videos have honed their craft over years of experience.
Jenny Nicholson’s older videos, which are shorter and have less razzle dazzle, are much lighter on production, but still a lot of time and effort. Hank Green does 2-3 vlogs per week on his channel and the vlogbrothers channel that I like a lot, which are relatively quick to make (they have to be, to come out so frequently). But Hank Green has this amazing radio host, podcaster, YouTube vlogger, TV personality-type charisma that is rare and also not a thing where you can say just go do that. If you think you can go do that, then definitely do it! But it’s not something just anyone can do, and no one can do it easily.
Another positive example from YouTube: Kirk Honda, a clinical psychology professor whose channel is called Psychology in Seattle. He does unbelievably good educational videos on mental health, psychology, personality disorders, therapy, abusive relationships, and so on, and the spoonful of sugar that makes the medicine go down is he’s often reacting to reality TV. But this is not a gimmick. He says that for ethical reasons, clinical psychology professors don’t have recordings of real people interacting that they can show to their students. Without real (or realistic) examples, it’s hard to convey what you’re talking about. Reality TV is a great teaching tool because these Hollywood studios have decided to do the arguably unethical thing and exploit people’s anguish for money, so he can just comment on it and use it to explain psychology concepts. These videos seem relatively quick to produce, but again you have the Hank Green problem: Kirk Honda has charisma, wisdom, eloquence, and charm, and not just anyone can do it, and no one can do it easily.
Podcasts are another medium that is incredibly popular, but also hard to do. Effective altruism already has the 80,000 Hours Podcast which is by and large a great success. The winning formula for shows like that is just booking fascinating people with fascinating things to say and letting them talk for a long time. And having good audio and video production, and having an interviewer who can lightly steer the fascinating person into saying more fascinating things. This is the easiest formula to replicate, but 80,000 Hours is already doing it. Not that I discourage people from starting new podcasts, or that I think the 80,000 Hours Podcast can’t be improved — my suggestion to them was to shake things up with guests who say things other than what people in EA are used to hearing, e.g., Richard Sutton, Yann LeCun, Jeff Hawkins, Edan Meyer. Toby Ord’s episode of the 80,000 Hours podcast was an amazing example of this. Toby Ord is as central to EA as it gets, but the cold water he’s dumped on AI scaling very much goes against most people in EA are saying about AI.
I’ve worked on 3 podcasts as hobby projects, and let me tell you, unless you’re already an expert, audio production is surprisingly hard to work out. If you’ve got a budget of ~$200 per episode (take this with a grain of salt and double-check my recollection/estimate), you can take the vast majority of this complexity away by renting a podcast recording studio for 1-2 hours. I think there are probably multiple spaces you can rent in any large city in developed countries. (This requires the people on the podcast are all physically in the studio. I don’t have a solution for making it easier to record podcasts remotely.) Editing the audio isn’t nearly as complicated and just takes time and work, but if you’ve got more money to spend, you can also hire an editor to take that complexity off your hands as well.
The right way to think about this is you’re trying to make good art. Easier said than done. However, I think it’s worth thinking about, and it’s worth trying for people who are willing to put in what it takes to give it a good shot. I just want to discourage people from thinking about making “content” or “social media content” as opposed to art.
Interesting!! Curious for any more detail on why you think this, if it's not too annoying to write out :)
Yep, that's one of the things I'd be super excited about!!
To that end, earlier this year I helped get started AI in Context, which has been heavily inspired by the awesome creators you mentioned above :)
It sure is time and resource-heavy to get the videos out (we've only managed two so far (working on the third!) even though we hired the first programme staff in early Feb), but my hope is that it's worth it — the reception has been broadly very positive :)