How my non-profit built the most affordable internet in the US

Hi, my name is Dylan Zajac. I started Computers 4 People when I was 15 years old, after learning that massive amounts of usable electronics were being exported as “e-waste” and burned in low-income countries, releasing toxic chemicals linked to serious health issues like respiratory illness and cancer. At the same time, people in my own community were being denied school and work opportunities simply because they didn’t have a computer. That gap, and the waste on both sides of it, felt deeply wrong.

I’ve been vegetarian since I was very young, vegan for a little over a year, and I strongly identify with Effective Altruism, especially the ideas of durability, counterfactual impact, and avoiding fragile systems. Computers 4 People began with donated laptops and no real funding, and over time grew into a nonprofit focused on refurbished computers, affordable internet, and basic digital skills.

One of the biggest shocks came when a major federal internet subsidy ended. Millions of low-income households lost connectivity almost overnight, including many we served. From an EA perspective, this felt like a textbook fragility failure: high impact while it exists, near-zero resilience when policy changes.

Rather than trying to fully replace that funding through donations, we ran a small experiment: build a boring, sustainable service people pay for and hard-link it to impact. That became Shield Internet, which follows a simple rule:

• every year of paid internet service funds one refurbished computer for someone in need
• the company is fully vegan, aligning with my personal ethics
• impact is tracked internally, without offsets or vague pledges

I don’t think this model generalizes everywhere, but it’s been a useful middle ground between grant-only charity and profit-maximizing approaches. I’m posting here to share the reasoning and to learn from others experimenting with durable, ethics-aligned ways to fund social goods.

(Not a pitch,  genuinely interested in feedback and counterarguments)