“Seeking personal assistant to influential philanthropist. Must be highly competent, discreet, and comfortable with polyamory and longtermism.”
That was a real job listing on an Effective Altruism (EA) job board
a few years back. It sounded like Black Mirror meets LinkedIn. But it captured something real: a movement that started with rigorous, data-driven giving had evolved into a swirl of fringe ethics, overconfidence, and tech-world vibes.
And then came the crash.
Intrigued? Read on. It’s a wild ride.
Thanks for reading Anthralytic’s Substack! Subscribe for free to receive new posts and support my work.
The FTX Fallout: When Altruism Became a Liability
Sam Bankman-Fried (SBF), a crypto billionaire and self-proclaimed EA, pledged to make billions so he could give it all away. And for a while, the EA world embraced him—funding AI safety think tanks, malaria prevention, and quirky projects through the FTX Foundation.
Then FTX imploded. Fraud, mismanagement, lawsuits. The money vanished, and with it, a big chunk of EA’s credibility.
Turns out “earn to give” doesn’t work so well when the “earn” part involves massive fraud.
And it raised a deeper question: How did a movement that prides itself on rationality and ethics get played so badly?
What EA Got Right (And Why It Matters)
I’ve spent nearly 15 years in international development, mostly on agriculture and MEL (monitoring, evaluation, and learning). So when EA showed up talking about “cost per life saved” and “evidence-based giving,” I was intrigued.
Here’s what they brought to the table:
Focus on outcomes. EA pushed philanthropy to ask: Does this actually work? And how much impact are we getting per dollar?
Willingness to fund unsexy things. Mosquito nets. Deworming pills. Direct cash transfers. All backed by evidence, even if they didn’t make for splashy press releases.
Radical transparency. GiveWell, an EA-aligned org, publishes detailed analyses, shows uncertainty, and updates recommendations when new data come in.
It was a rare thing in the social sector: money following evidence.
That felt like fresh air compared to the vague “we empower communities” language I’ve seen (and written) in countless reports.
Where It Went Off the Rails
And then came the weirdness. A few shifts in the movement took EA from interesting to yikes:
Longtermism. EA’s loudest voices started focusing not on global poverty or health—but on hypothetical future risks. Like… AI turning us into paperclips. Seriously. That was an actual discussion.
Utilitarian overreach. In trying to do the most good, EA started to justify almost anything—including risky or unethical behavior—if the outcome might be positive.
Groupthink. A small circle of white, male philosophers and tech billionaires decided what counted as “effective.” That’s not a great setup for spotting blind spots and therefore inefficiencies.
As Nate Silver put it in his latest book, On the Edge1, EA became “too trusting, too selfless, and too reckless.” They were optimizing for good, but forgot the human side of doing good—like trust, accountability, and ethics.
They fell in love with the spreadsheet and forgot to reality-check the formula.
What We Should Keep
Despite the fallout, I think we’d be foolish to ditch everything EA stood for. A few principles still hold up:
Impact over optics. EA’s donors funded what worked—not just what looked good. That’s a mindset we need more of in development and philanthropy.
Question default settings. Why do we assume helping nearby is better than helping effectively? EA challenged that assumption, and it’s worth rethinking.
Be ready to pivot. EA organizations are unusually open to changing their minds. That’s rare. And refreshing.
But we need to pair those with the things EA often forgot:
Humility. You don’t know everything. Especially if you’ve never set foot in the communities you’re funding.
Diverse perspectives. No movement thrives when everyone looks, thinks, and models impact the same way. We can not be optimally effective or efficient if we fail to see our own blind spots.
Ethical guardrails. If your math says “fraud is fine,” your math needs a rewrite.
What This Means for the Rest of Us
At Anthralytic, we try to bring these lessons into the real world. We build tools that help nonprofits figure out what’s working, when to pivot, and how to track results—without losing the plot or the people.
We’ve seen over and over that:
Data is only powerful when it’s paired with lived experience.
You can optimize all day—but if you’re not asking the right questions, or listening to the folks affected, you’re just creating elegant nonsense.
TL;DR
Effective Altruism brought some much-needed rigor to the social sector.
Then it spiraled into sci-fi ethics, groupthink, and trusting the wrong billionaire.
But we should keep what worked: impact focus, transparency, and the willingness to rethink.
And pair it with what EA lacked: humility, diversity, and common sense.
Maybe we don’t need to “maximize expected utility” to the tenth decimal point. Maybe we just need to do better by thinking harder—and staying human.
