AI Will Make the Digital Divide More Expensive

My son may inherit a world increasingly split by who can use AI well and who can afford the tools, time, and training to learn it. That divide is not inevitable, but the early pricing and education signals point in that direction.

The Checkout Line Tells You Everything

Some people walk through Amazon Go without stopping. Their digital profiles and payment methods handle everything. Others wait in traditional checkout lines, struggling with card readers, locked out of conveniences that the digitally privileged take for granted.

I expect a similar split with AI-enhanced services. People with access get faster ways through a system. People without it may be left with an interface companies have less incentive to improve.

The Poverty Premium

The credit system charges more to people who have less money. Higher interest rates. Rejected mortgage applications. Worse credit terms. Economists call it the “poverty premium.”

For the affluent, favorable mortgages and investment access turn debt into wealth. For the economically vulnerable, high interest rates, fees, and penalties make debt extract wealth. Financial stratification amplifies class divisions instead of flattening them.

Buying Time Is Privilege

When someone orders delivery instead of shopping, they buy back time by paying for someone else’s labor. That transaction is not inherently exploitative; the pay, protections, and bargaining power determine whether it is a fair exchange. The privilege is having the option to convert money into time at all.

That labor is becoming more fragmented, not less. Gig platforms after Prop 22 shows what happens when driving, delivery, and small AI tasks all compete for the same worker’s time.

Two Futures Are Forming

In one, people become fluent in AI collaboration. They access the latest models, develop prompting expertise, and delegate routine tasks to focus on work that machines can’t do. AI literacy becomes power.

In the other, people compete against AI systems instead of working with them. Without access to advanced tools or training, they struggle to show value in a workforce where machines handle increasingly complex tasks. They service the AI economy rather than direct it.

I do not know whether public education will adapt quickly enough. I do know that families with money are already able to buy subscriptions, devices, and time to experiment. AI subscription pricing makes that advantage visible: capability, context, and usable quotas increasingly sit on different price tiers.

If access and AI literacy remain things families have to buy for themselves, my son will not inherit a neutral tool. He will inherit another mechanism that can turn existing privilege into more privilege.