Podcast Episode: AI And The Future Of Work

Pip: Hemant Pandey writes about the future of work, and if you've been quietly hoping AI would just take the boring parts and leave everything else intact — well, he has some gently devastating news for you.

Mara: The posts we're covering today map two connected territories: what AI is actually doing to jobs and economic ladders, and what human roles might look like once the dust settles. Let's start with the employment picture itself.

The job isn't gone — it's just worth less

Pip: The AI-and-jobs debate has mostly been asking which occupations survive. The more unsettling question is whether surviving an occupation is even enough anymore.

Mara: The post "New World and AI Taking Jobs Scenario" sets this up precisely. The setup is that AI doesn't need to eliminate a job title to hollow it out — and here's the line that captures it: "Your job still exists, but society needs fewer hours of it, fewer people doing it, and perhaps eventually fewer people paid as much for doing it."

Pip: So the first sign of disruption isn't a pink slip. It's slower wage growth, fewer entry-level openings, and a quiet raise in what's expected from whoever's left.

Mara: That matches the evidence cited. Apollo Global Management research found real wage growth in AI-exposed occupations lagged less-exposed ones by 6.7 percentage points after 2023, with no broad collapse in employment. Anthropic's March 2026 analysis found suggestive evidence of slower hiring of younger workers in exposed fields. The job survives; the ladder gets shorter.

Pip: And "When the Ladder Disappears" names exactly that problem — not unemployment, but the disappearance of the entry-level rungs that were always the training ground.

Mara: Right. That post frames it as a building with ten floors. Historically, technology removed work from the bottom and created new work above it. AI can attack several floors simultaneously, which means ordinary people may be expected to enter the economy near the top — as entrepreneurs, decision-makers, or people with rare capabilities — without ever having climbed there.

Pip: Which is a bit like removing the first nine floors of a building and telling people the penthouse is still available.

Mara: The third piece, "The Era of Universal Employment," proposes a structural response to exactly this problem. It argues that almost everyone could have a pathway to participate in producing economically useful intelligence — not through conventional jobs, but through contributions: corrections, teaching strategies, domain expertise, edge cases that improve AI systems.

Pip: The phrase used is a "Knowledge Dividend" — closer to royalties than to a paycheck.

Mara: And the post is careful to call this a conceptual framework, not a claim that the mechanisms already exist. The harder question it leaves open is what happens to an economy built around human labor when human labor is no longer the primary input.

Pip: From there, the question shifts to what humans actually do inside that economy.

What humans contribute when machines do the work

Pip: Once AI handles the output, what's the distinctly human part of the transaction? "The Evolution of Humans in the New Economy" argues the answer is cognition — not content.

Mara: The post traces the arc from consumer to creator to something newer. The key line: "When intelligence becomes abundant, judgment may become scarce." The user who corrects a model, supplies context, and decides which answer is actually useful is doing something the post calls cognitive contribution — work that happens inside the interaction, invisible as labor but potentially significant as value.

Pip: So the question isn't just what AI can do. It's who controls what AI should be used for — and whether that authority is the last thing left that can't be automated.

Mara: That's where both posts land. The scarcest resource may not be intelligence. It may be the agency to decide what intelligence is for.

Pip: The ladder, the rungs, the wage gap, the knowledge dividend — these are all different ways of asking the same question: who benefits when machines get smarter?

Mara: And whether the answer is "everyone" or "whoever owns the system" may depend on choices that haven't been made yet. More on those choices next time.