Hi Ben, Thank you for the thoughtful reply. Super great to see a greater focus on community culture in your plans for 2020. You are always 2 steps ahead :-)
That said I disagree with most of what you wrote.
Most of your reply talks about communications hurdles. I don’t think these pose the barrier you think they pose. In face the opposite, I think the current approach makes communications and mistrust issues worse.
You talk about the challenge of being open about your prioritisation and also open to giving advice across causes, risks of appearing to bait and switch, transparency Vs demoralising. All of these issues can be overcome, and have been overcome by others in the effective altruism community and elsewhere. Most local community organisers and CEA staff have a view on what cause they care the most about yet still mange an impartial community and impartial events. Most civil servants have political views but still provide impartial advice to Ministers. Solutions involve separating your priotisation from your impartial advice, having a strong internal culture of impartiality, being open about your aims and views, being guided by community interests, etc. This is certainly not always easy (hence why I had so many conversations about how to do this well) but it can be done.
I say the current approach makes these problems worse. Firstly thinking back to my time focused on local community building (see examples above) it appeared to me that 80000 Hours had broken some of the bonds of trust that should exist between 80000 Hours and its readership. It seems clear that 80000 Hours was doing something wrong and that more impartiality would be useful. (Although take this with a pinch of salt as I have been less in this space for a few years now). Secondly it seems surprising to me that you think the best communications approach for the effective altruism community is to have multiple organisations in this space for different causes with 80000 Hours being an odd mix of everything and future focused. A single central organisation with a broader remit would be much clearer. (Maybe something like franchising out the 80000 Hours brand to these other organisations if you trust them could solve this.)
I fully recognise there are some very difficult trade-offs here: there is a huge value to doing one thing really well, costs of growing a team to quickly to delve into more areas, costs of having lower impact on the causes you care about, costs of switching strategy, etc.
Separately to the above I expect that I would place a much stronger emphasis than you on epistemic humility and have more uncertainty than you about the value of different causes and I imagine this pushes me towards a more inclusive approach.
It’s true a lot of my reply was about the communications challenges, but that’s because they’re harder to explain, I didn’t mean to imply they were the most significant reason for our strategy. The first and probably most important tradeoff I mentioned was:
Going broader makes the community more inclusive and so potentially larger, but at the cost of more of our effort going on areas that we think have far lower expected impact.
This is a huge topic, so I’m not going to be able to debate it here, but I wanted to flag this is probably the key driver of ours views (e.g. the question of how much causes differ in effectiveness, as you mention at the end).
However, I did want to respond to your point that you think it’s clear 80k was doing something wrong in the past, and in particular that we had broken bonds of trust with our readership. I’m sad to hear you think our lack of ‘cause impartiality’ has betrayed our readers’ trust. 80k’s key mission and responsibility is to improve the lives of others as much as possible. This requires us to prioritize between causes.
I actually think that being clear about what we’re prioritizing is an important part of fulfilling our duty to readers, and more importantly, it’s central to our duty to those whose lives we’re seeking to improve.
You might think that despite that being our overall goal, the way we should be achieving it is supporting the EA community. That could mean prioritising content according to average views across the whole community rather than using our own judgement or the judgement of those working in global priorities research (as well as doing the other things you mention, which seem like good suggestions). However, I don’t think that’s the best way to pick causes, and our readership also extends beyond the EA community, so we would still run into the problem with the wider audience.
I think that problems mainly arise if you mix these two strategies, rather than picking one and being clear about it.
I agree that in the past we sometimes portrayed ourselves as a more general source of EA careers advice than we were. I regret if this led people to be disappointed, or slowed down the creation of alternative sources of advice. In recent years we’ve been clearer about our role, and the fact that our site is about our priorities. We want our readers to be able to decide for themselves whether it’s useful to them.
We’ve found the impression of our site as more general to be hard to shift among the broader community, unfortunately. That’s why I wanted to write a post here to lay out our views and priorities clearly. I hope this will make it easier for community organisers to understand what 80k can provide over the next couple of years.
I appreciate it would be simpler if 80k could be a one-stop shop for EA careers advice, though we have worked around this problem in other areas e.g. GiveWell is the largest ‘where to donate’ org, but it only covers global poverty.
On the solutions for the EA community, I should have been clearer that I think where we most need multiple orgs is when it comes to one-on-one support and specialised content. We simply don’t have capacity to cover every cause area, and won’t for some time. We still provide general purpose advice (e.g. on principles like career capital), and as noted, we’re also often happy to link to or send people to other sources of advice, so can still act as a clearinghouse.
I also have the aspiration that 80k gets broader over time, so I hope we cover some of these gaps ourselves. However, the process will be faster if there are other groups involved.
“I actually think that being clear about what we’re prioritizing is an important part of fulfilling our duty to readers [...] You might think that despite that being our overall goal, the way we should be achieving it is supporting the EA community.”
My reading of this disagreement is a bit different.
A few years ago, when I encountered 80,000 Hours’ career guide, it seemed like your goal was “to help readers to help others by creating cause-neutral way to think systematically about their career”.
In 2018 it seemed more like 80k’s goal was “to recruit readers to work in AI and ops”.
I don’t know if your goals changed, or if the way you communicated them changed, but the first goal seems better.
Hi Ben, I think you are correct that the main difference in our views is likely to be the trade-off between breadth/inclusivity verses expected impact in key areas. I think you are also correct that this is not a topic that either of us could do justice in this thread (I am not sure I could truly do it justice in any context without a lot of work, although always happy to try). And ultimately my initial disappointment may just be from this angle.
I do think historically 80K has struggled more in communicating its priorities to the EA community than others (CEA / GiveWell / etc) and it seems like you recognise this has been a challenge. I think perhaps it was overly harsh of me to say that 80K was “clearly doing something wrong”. I was focusing only on the communications front. Maybe the problems were unavoidable or the past decisions made were the net best decisions given various trade-offs. For example maybe the issues I pointed to were just artifacts of 80K at the time transitioning its messaging from more of a “general source of EA careers advice” to more of cause focused approach. (It is still unclear to me if this is a messaging shift or a strategy shift). Always getting messaging spot on is super difficult and time consuming.
Unfortunately, I am not sure my thoughts here have lead to much that is concretely useful (but thank you for engaging). I guess if I had to summarise some key points I would say: I am super in favour of transparency about priorities (and in that regard this whole post is great); if you are focusing more on your effect on the effective altruism movement then local community organisers might have useful insights (+CEA ect have useful expertise); if 80k gets broader over time that would be exciting to me; I know I have been critical but I am really impressed by how successful you have made 80k.
Hi Ben, Thank you for the thoughtful reply. Super great to see a greater focus on community culture in your plans for 2020. You are always 2 steps ahead :-)
That said I disagree with most of what you wrote.
Most of your reply talks about communications hurdles. I don’t think these pose the barrier you think they pose. In face the opposite, I think the current approach makes communications and mistrust issues worse.
You talk about the challenge of being open about your prioritisation and also open to giving advice across causes, risks of appearing to bait and switch, transparency Vs demoralising. All of these issues can be overcome, and have been overcome by others in the effective altruism community and elsewhere. Most local community organisers and CEA staff have a view on what cause they care the most about yet still mange an impartial community and impartial events. Most civil servants have political views but still provide impartial advice to Ministers. Solutions involve separating your priotisation from your impartial advice, having a strong internal culture of impartiality, being open about your aims and views, being guided by community interests, etc. This is certainly not always easy (hence why I had so many conversations about how to do this well) but it can be done.
I say the current approach makes these problems worse. Firstly thinking back to my time focused on local community building (see examples above) it appeared to me that 80000 Hours had broken some of the bonds of trust that should exist between 80000 Hours and its readership. It seems clear that 80000 Hours was doing something wrong and that more impartiality would be useful. (Although take this with a pinch of salt as I have been less in this space for a few years now). Secondly it seems surprising to me that you think the best communications approach for the effective altruism community is to have multiple organisations in this space for different causes with 80000 Hours being an odd mix of everything and future focused. A single central organisation with a broader remit would be much clearer. (Maybe something like franchising out the 80000 Hours brand to these other organisations if you trust them could solve this.)
I fully recognise there are some very difficult trade-offs here: there is a huge value to doing one thing really well, costs of growing a team to quickly to delve into more areas, costs of having lower impact on the causes you care about, costs of switching strategy, etc.
Separately to the above I expect that I would place a much stronger emphasis than you on epistemic humility and have more uncertainty than you about the value of different causes and I imagine this pushes me towards a more inclusive approach.
It’s true a lot of my reply was about the communications challenges, but that’s because they’re harder to explain, I didn’t mean to imply they were the most significant reason for our strategy. The first and probably most important tradeoff I mentioned was:
This is a huge topic, so I’m not going to be able to debate it here, but I wanted to flag this is probably the key driver of ours views (e.g. the question of how much causes differ in effectiveness, as you mention at the end).
However, I did want to respond to your point that you think it’s clear 80k was doing something wrong in the past, and in particular that we had broken bonds of trust with our readership. I’m sad to hear you think our lack of ‘cause impartiality’ has betrayed our readers’ trust. 80k’s key mission and responsibility is to improve the lives of others as much as possible. This requires us to prioritize between causes.
I actually think that being clear about what we’re prioritizing is an important part of fulfilling our duty to readers, and more importantly, it’s central to our duty to those whose lives we’re seeking to improve.
You might think that despite that being our overall goal, the way we should be achieving it is supporting the EA community. That could mean prioritising content according to average views across the whole community rather than using our own judgement or the judgement of those working in global priorities research (as well as doing the other things you mention, which seem like good suggestions). However, I don’t think that’s the best way to pick causes, and our readership also extends beyond the EA community, so we would still run into the problem with the wider audience.
I think that problems mainly arise if you mix these two strategies, rather than picking one and being clear about it.
I agree that in the past we sometimes portrayed ourselves as a more general source of EA careers advice than we were. I regret if this led people to be disappointed, or slowed down the creation of alternative sources of advice. In recent years we’ve been clearer about our role, and the fact that our site is about our priorities. We want our readers to be able to decide for themselves whether it’s useful to them.
We’ve found the impression of our site as more general to be hard to shift among the broader community, unfortunately. That’s why I wanted to write a post here to lay out our views and priorities clearly. I hope this will make it easier for community organisers to understand what 80k can provide over the next couple of years.
I appreciate it would be simpler if 80k could be a one-stop shop for EA careers advice, though we have worked around this problem in other areas e.g. GiveWell is the largest ‘where to donate’ org, but it only covers global poverty.
On the solutions for the EA community, I should have been clearer that I think where we most need multiple orgs is when it comes to one-on-one support and specialised content. We simply don’t have capacity to cover every cause area, and won’t for some time. We still provide general purpose advice (e.g. on principles like career capital), and as noted, we’re also often happy to link to or send people to other sources of advice, so can still act as a clearinghouse.
I also have the aspiration that 80k gets broader over time, so I hope we cover some of these gaps ourselves. However, the process will be faster if there are other groups involved.
“I actually think that being clear about what we’re prioritizing is an important part of fulfilling our duty to readers [...] You might think that despite that being our overall goal, the way we should be achieving it is supporting the EA community.”
My reading of this disagreement is a bit different.
A few years ago, when I encountered 80,000 Hours’ career guide, it seemed like your goal was “to help readers to help others by creating cause-neutral way to think systematically about their career”.
In 2018 it seemed more like 80k’s goal was “to recruit readers to work in AI and ops”.
I don’t know if your goals changed, or if the way you communicated them changed, but the first goal seems better.
Hi Ben, I think you are correct that the main difference in our views is likely to be the trade-off between breadth/inclusivity verses expected impact in key areas. I think you are also correct that this is not a topic that either of us could do justice in this thread (I am not sure I could truly do it justice in any context without a lot of work, although always happy to try). And ultimately my initial disappointment may just be from this angle.
I do think historically 80K has struggled more in communicating its priorities to the EA community than others (CEA / GiveWell / etc) and it seems like you recognise this has been a challenge. I think perhaps it was overly harsh of me to say that 80K was “clearly doing something wrong”. I was focusing only on the communications front. Maybe the problems were unavoidable or the past decisions made were the net best decisions given various trade-offs. For example maybe the issues I pointed to were just artifacts of 80K at the time transitioning its messaging from more of a “general source of EA careers advice” to more of cause focused approach. (It is still unclear to me if this is a messaging shift or a strategy shift). Always getting messaging spot on is super difficult and time consuming.
Unfortunately, I am not sure my thoughts here have lead to much that is concretely useful (but thank you for engaging). I guess if I had to summarise some key points I would say: I am super in favour of transparency about priorities (and in that regard this whole post is great); if you are focusing more on your effect on the effective altruism movement then local community organisers might have useful insights (+CEA ect have useful expertise); if 80k gets broader over time that would be exciting to me; I know I have been critical but I am really impressed by how successful you have made 80k.