“Don’t burn out. Seriously, don’t do it,” is a phrase that comes up in effective altruism spaces. It’s written in blog posts, said during lunch-time stories, and often, it’s prompted by someone burning out.
Unfortunately, I don’t think warning flags actually accomplish that much. Not burning out isn't really a mantra—it's an active practice requiring more than individual willpower. As many have observed, burnout is often structural, though individual factors certainly play a role.
Why it’s hard to prevent burnout
1. “I won’t know how burnout feels until I experience it (and I don’t want to leave the kitchen)”
Forgive the infantilising metaphor, but some kids will want to touch the stove top at least once, just to know for themselves if it truly hurts. And as much as their parents might pull them back, exasperated, both parties are simply acting on the information they have.
Now imagine if: to accomplish your goals, you have to remain within some distance of the stove. Everyone around you keeps getting really close to it, while also telling you not to touch it. So you get closer and closer and you start to wonder, “is this too close? Should I pull back?” But alas, if you pull back, you might have to leave the kitchen. And you don’t want to. And how bad does the stove really hurt, anyways?
2. “Trade-offs are real”
“If I stop working as hard, I’ll get less done.”
This might be true—or it might not. It depends. You could discover that a sustainable workload actually maintains your productivity level while improving work quality. It could also prevent burnout and increase your overall happiness and long-term impact. But there's no guarantee. A more sustainable approach might indeed mean accomplishing less in the short term, maybe even less overall.
This is a tradeoff that everyone is allowed to evaluate for themselves.
3. “I’m genuinely scared of global suffering and AI risk”
I hear you. By its nature, effective altruism can lean toward a dissolution of self: it emphasizes scale, impact, and often touches on notions of moral obligation and neutrality. It isn't designed to encourage individuation, which isn't inherently bad or unique. Many worthwhile philosophies encourage individual sacrifice for community or causes. It is one of the things I love about EA, which says: countless people are suffering preventably, vast potential exists in our future, high-stakes problems demand attention right now, and you can concretely do something about it.
Many take this idea very seriously and they are surrounded by people who also take it seriously. And so, they work hard. I see nothing inherently wrong with this; there's beauty in witnessing people who are willing to make personal sacrifices, not for what I might consider “empty” achievement, but for something meaningful.
"Don't burn out" is not a value—it's an action
Preventing burnout is challenging for individuals naturally susceptible to it through some combination of values, environment, and experiences. For some, no number of blog posts insisting "don't!" will help them move their hand away from the stove.
So what does this mean?
Organisations can (and should) protect their employees
Not burning out is partly embedded in organisational culture and policies.
On the cultural level: How close are most people to the stove? For example, does your organisation generally model and encourage:
Taking time off
Reducing project ambition or scope in favour of sustainable workloads
Sticking to set working hours
Communicating and respecting boundaries
Not excessively celebrating “hero” culture (where employees take on great personal cost to achieve a goal)
etc.
On the policy level: Does anyone pull people away from the stove? Does your organisation have:
Systems for meaningful check-ins between managers and reports
Avenues for reports to check in outside their direct manager (skip-level meetings, HR resources)
Concrete contingency plans (what happens when someone says "this is too much"?)
Individual choice still matters, but organisations shouldn’t have the same freedom
The world is complicated. Everyone has their own views, priorities, and values, and ultimately, people must make their own choices. But at managerial or institutional levels, those individual choices affect others.
That's when I feel comfortable insisting that mental wellness be taken seriously, and I don’t think that’s accomplished by telling individuals they shouldn’t burnout (though to be clear, I find it very brave and inspiring when people share their personal experiences. I think this is helpful for others who might be struggling, but I don’t think it’s the solution). And for this reason, I'm pretty unhappy every time I hear a rumour about another AI safety organisation overworking employees or dismissing burnout concerns—not because I can't understand their perspective, but because organisations aren't individuals. Organisations must follow both legal and ethical standards, which means they have a responsibility to care about employee health.
In a movement dedicated to doing the most good, we cannot ignore the well-being of those doing the work. The stove will always be hot. And the answer can’t simply be, “don’t touch it.”
This was a good post! It's a compelling read and also importantly correct
The only reliable check for burnout I've found is thinking of whether I expect enough success from my project as will justify the effort I put in. This seems much more predictive than hours worked, or how burned out you feel (often NOT burned out just before you collapse, as you struggle to get something over the finish line). You should be getting more from a project than what you put in, not vice versa.
In that context, I don't think of "sustainability" as a reliable, long-term sustainability, but more along the lines of "will the amount of gut-expected success from this project actually make me feel energy-positive afterward?" A lot of people seem to only be tracking abstract expected impact in a way that they do not believe on a gut level, and I think being real about your feeling of "success" here can really improve matters. Obviously success can include things like working with great people on a great bet that doesn't pan out, but it's easy to fool yourself that a bet is worth it if you're not serious about your gut. Lots of good people and good bets will still not reach your gut's bar.