What should I decide first: the type of organization to support, or the outcome I want to achieve?

Start with the outcome you want to have on the world, not with picking which kinds of organizations to back. Deciding upfront on categories (nonprofits vs. political giving) is the most common practice, but Bahat argues this approach is backward and can prematurely tie your hands. More than one big foundation has wished their donor hadn't bonded the money to strictly nonprofit giving.

How much control should I keep over how my money is spent?

Bahat suggests learning from startup investors, who generally trust founders to make the best choices about how to deploy capital without imposing restrictions. In contrast, nonprofits often receive "restricted giving," where donors insist on particular uses. He notes this is problematic because donors often know less about nonprofits than they do about startups, yet are far more controlling in the former case.

Can I learn by starting small?

Yes. Bahat advises beginning at whatever scale feels low-risk, since some learning happens by doing. However, he cites Zuckerberg's early $100M gift to Newark education—made before creating the Chan Zuckerberg Initiative—as a cautionary tale that much of this work is too important to A/B test, and that early giving can sometimes go awry.

Hacker NewsJul 26, 2026Jul 27, 2026, 01:01 JST

AI Workers With Newfound Wealth: A Guide to Giving It Away

Send on LINE

3 Key Points

What happened
Roy Bahat, writing for those in AI who want to use their money to help the world, offers a framework for effective charitable giving. He emphasizes starting with the outcome you want to achieve, not with organizational categories (nonprofits vs. political giving), and warns against prematurely restricting how money is used.
Why it matters
People in AI are under unusual scrutiny—even small philanthropic moves draw attention. The piece argues that many wealthy donors inadvertently limit their impact by bonding money to rigid categories upfront, and that AI workers can learn from both startup investing (which trusts founders) and nonprofit practice (which often micromanages). Understanding your theory of change and considering participatory grantmaking (sharing decision-making with those affected by the issues) can unlock better outcomes.
What to watch
Bahat cites cautionary examples—like Zuckerberg's early $100M Newark education gift, made before he created the Chan Zuckerberg Initiative—as a reminder that learning by doing at smaller scale first can prevent costly mistakes. He also notes that giving money away is only part of the puzzle; examining what privileges and power you are willing to sacrifice matters equally.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Get these free every morning

Context & Analysis

Bahat's essay frames charitable giving not as a logistics problem (how to distribute money across buckets) but as a strategic one rooted in what you believe needs to happen in the world. This reframing matters especially for AI workers because they inhabit a moment of unprecedented wealth creation and, simultaneously, unusual public attention. A small makerspace gift from an OpenAI engineer made local news—a sign that donors in the field cannot give quietly. That visibility creates both opportunity and obligation to think carefully.

The essay also draws a subtle contrast between two models of philanthropy. Startup investing operates on trust and founder autonomy; nonprofit philanthropy, historically, operates on donor control ("restricted giving"). Bahat suggests this asymmetry is misguided—that the nonprofit sector, where donors often know less, is where they most insist on control. His implicit argument is that ceding power to those who know the work best—and to those most affected by the problems—produces better outcomes than top-down micromanagement.

Central to his framing is the idea that giving money is only one piece of the work. He quotes Lilla Watson: "If you have come here to help me you are wasting your time, but if you have come because your liberation is bound up with mine, then let us work together." This suggests that the deeper question is not how much to give, but what privileges and power each donor is willing to examine and sacrifice. For AI workers with accelerated wealth, that examination may be the harder and more consequential work.

FAQ

What should I decide first: the type of organization to support, or the outcome I want to achieve?: Start with the outcome you want to have on the world, not with picking which kinds of organizations to back. Deciding upfront on categories (nonprofits vs. political giving) is the most common practice, but Bahat argues this approach is backward and can prematurely tie your hands. More than one big foundation has wished their donor hadn't bonded the money to strictly nonprofit giving.
How much control should I keep over how my money is spent?: Bahat suggests learning from startup investors, who generally trust founders to make the best choices about how to deploy capital without imposing restrictions. In contrast, nonprofits often receive "restricted giving," where donors insist on particular uses. He notes this is problematic because donors often know less about nonprofits than they do about startups, yet are far more controlling in the former case.
Can I learn by starting small?: Yes. Bahat advises beginning at whatever scale feels low-risk, since some learning happens by doing. However, he cites Zuckerberg's early $100M gift to Newark education—made before creating the Chan Zuckerberg Initiative—as a cautionary tale that much of this work is too important to A/B test, and that early giving can sometimes go awry.

Hacker NewsRead Original Article

Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Get it free

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Next articleSalesforce edges ServiceNow as cheaper AI play despite slower growth