AI Gains Favor the Rich; Workers Deserve Strong Protections

They’re right about one thing: AI-driven productivity will concentrate returns unless someone rigs the rules. Fortune’s piece is blunt about gains piling up at the top and jobs getting wiped out, while the IMF chief reaches for a “silver lining” for low-wage workers. That’s a useful alarm bell. It’s also nowhere near enough.

Look, productivity is not a public utility — it doesn’t trickle, it gets allocated.

Rich as Default — You Get What You Pay For

In any firm or economy, owners and capital get first claim on the surplus. Wages, work design, and job security come later, if at all. When companies deploy AI to automate tasks and boost output, the default setting is simple: CEOs and shareholders keep the upside unless institutions, regulation, or bargaining power force a different split.

So yes, “AI makes the rich richer” is a fair description. The missing move in the Fortune piece is prescriptive: how to turn that extra productivity into shared gains instead of just nicer earnings calls.

Jobs aren’t a single bucket. Some roles will disappear, others will morph, and new ones will show up in odd corners — not always in the same regions, industries, or pay bands. Low-wage workers can benefit from AI, but only if businesses and policymakers deliberately redesign work around human strengths: relationship-heavy service, on-site problem solving, local knowledge. And then tie pay and progression to those strengths, not just to how well people babysit software.

Without explicit redistribution mechanisms — higher minimum wages, sectoral wage floors, portable benefits, or corporate profit-sharing — the trajectory is obvious: concentration, then resentment.

I spent years in operations at a large corporation watching the same movie. We’d run “efficiency initiatives” that cut cycle time, trimmed headcount, and freed up cash. The celebration was always about margin expansion, rarely about front-line pay or workload sanity. Efficiency is morally neutral; what you do with the extra margin is not.

The IMF’s “Silver Lining” Is a Recipe Without Ingredients

The IMF chief’s optimism — that AI might carry a silver lining for low-wage workers — is plausible only if it’s attached to actual tools. The Fortune article nods at that hope, then leaves it hanging.

If the idea is that automation pressure on higher-paid roles will make low-wage labor relatively more attractive, that’s an economic story with a lot of “ifs” baked in. It’s also a story that often breaks down by geography: a call center worker in one country doesn’t automatically benefit when an office job in another gets automated away.

If the hope is retraining and “skills upgrades,” that’s not a slide deck, it’s an infrastructure problem. You need funded programs, fast curriculum updates, and employers willing to hire based on new certificates instead of old pedigrees. None of that appears just because some software got smarter.

And if the implied route is social safety nets funded by higher taxes on AI-enhanced profits, then we’re back in the world of tax design and political hardball. Rents don’t tax themselves.

The central gap in the “silver lining” framing is this: the upside is a policy bundle, not a natural side effect of new tools. Pointing to notional benefits without specifying taxes, bargaining reforms, or direct wage interventions is a bright headline on top of a very thin mattress.

Three practical levers deserve more airtime than feel-good speculation:

• Tie firm incentives to sharing gains. Tax credits or regulatory relief linked to verifiable wage growth, mandated profit-sharing, or equity pools that include line workers instead of just executives.
• Raise bargaining power where AI will bite hardest. Sectoral wage standards, portable benefits, and simple organizing rules in logistics, retail, and basic services — the places most likely to see “AI-assisted” speedups turned into harder, faster shifts.
• Fund rapid, employer-partnered reskilling. Training aimed at roles that complement AI — supervision, exception handling, complex customer care — rather than funneling everyone into generic coding bootcamps.

These are not slogans. They’re operational levers you can write into law or contracts.

Speed is Real — But So Is Excuse-Making

The skeptics’ line is that AI is moving too fast for policy to keep up; industries shift in months while legislatures crawl, so trying to steer distribution is futile.

Give me a break.

Speed is a constraint, not a permission slip to shrug. Fast change calls for simple, blunt tools: automatic surtaxes on windfall profits tied to documented productivity spikes, temporary wage supplements in regions with heavy AI-related layoffs, emergency retraining vouchers keyed to specific industries rather than vague “future skills.” None of those are pretty long-term architecture, but they buy time while better frameworks are built.

We’ve done versions of this before. When ATMs rolled out, the prediction was mass bank teller extinction. What actually happened was messier: teller roles shifted toward sales and customer service, branch networks evolved, and employment didn’t collapse overnight. The difference now is scale and speed. AI touches every sector, not one niche, which makes the distributional question far more dangerous to ignore.

The Fortune piece correctly flags inequality as the headline risk. But two other consequences should worry us just as much: regional decline if AI investment clusters in a few affluent hubs and bypasses everywhere else, and a gradual erosion of social trust when people see productivity gains privatized while the fallout — unemployment, retraining, mental health strain — gets billed to the public.

You can’t keep promising a “silver lining” to low-wage workers while treating the policy toolkit as an afterthought. If AI is going to make the rich richer, as the article warns, then the real story will be whether anyone bothers to fight over how that new wealth gets carved up — and who’s willing to turn vague concern into hard, enforceable rules.