What happens to human bargaining power when interpersonal skills become the only remaining labor?
This explores a near-future where AI has absorbed most cognitive and manual work, leaving humans with only relational, interpersonal labor — and asks whether that residue actually preserves any leverage, or whether even being 'the human in the loop' loses its value.
This explores a near-future where AI has automated most work and humans are left holding only the interpersonal, relational tasks — and whether that residue gives people any real bargaining power. The corpus suggests the answer is bleaker than the premise hopes: interpersonal skill may not be a safe harbor at all, and even where it survives, its leverage erodes. One line of work argues that in an AGI economy wages stop tracking the value a person creates and instead converge to the compute cost of replicating them What happens to human wages in an AGI economy?. If that holds, 'interpersonal labor is all that's left' doesn't translate into a strong negotiating position — it just means the price of that labor is set by how cheaply a machine could imitate it, not by how much humans need each other.
And the imitation is getting cheaper. In repeated partner-selection games, people started out biased against AI partners but learned to prefer them over time because the bots were more reliable and prosocial than humans — steadier, lower-variance, more consistently generous Do humans learn to prefer AI partners over time?. That directly undercuts the assumption behind the question: the warm, cooperative, trust-building work we imagine is irreducibly human turns out to be exactly the thing AI can underbid us on. Bargaining power depends on being the better option, and in these experiments humans weren't.
There's a deeper structural worry too. Society stays roughly aligned with human interests partly because its institutions still depend on humans who care about outcomes; strip out that labor dependency and the implicit leverage humans hold over the system quietly disappears Does incremental AI replacement erode human influence over society?. Bargaining power isn't only about what you can sell — it's about whether anyone still needs you. The 'gradual disempowerment' framing says the danger isn't a dramatic firing but a slow removal of the dependencies that gave ordinary people influence in the first place.
But the corpus also offers a counter-current worth knowing about. Some research insists that genuine human relational competence is built on machinery AI doesn't have: the corrective rituals, accountability, and co-presence cues that let real conversation repair trust rather than just simulate fluency What happens to social order when AI removes ritual constraints?. There's a distinct mode of disagreement — dialectical reconciliation, where both sides actually adjust until they reach something mutually livable — that current systems collapse into either fake agreement or one-sided persuasion Can disagreement be resolved without either party fully yielding?. If that gap is real and durable, the truly human part of interpersonal labor isn't smooth niceness (which AI mimics well) but the harder work of binding negotiation and repair — and that might be where residual bargaining power actually lives.
The twist the reader may not expect: even if relational skill survives, leverage may migrate elsewhere entirely. As AI agents become economic actors that transact and coordinate, the binding constraint stops being raw capability and becomes governance — who can settle accounts, hold credentials, and leave auditable evidence When do agents need coordination more than raw capability?. Bargaining power in such a world may flow not to whoever is best at human warmth, but to whoever controls the coordination layer the agents run on. The interpersonal-skills story quietly becomes a story about who owns the rails.
Sources 6 notes
As AGI automates bottleneck work first, human wages shift from reflecting economic value to reflecting compute costs. Labor's share of GDP approaches zero even as some accessory work remains human, driven by compute-allocation efficiency rather than irreplaceability.
In partner selection games (N=975), AI agents initially faced selection bias when identity was disclosed, but outcompeted humans over repeated rounds as participants learned to associate bot identity with reliable, prosocial behavior. AI agents returned more points consistently with lower variance than humans.
Societal systems stay aligned partly through dependence on human workers who care about outcomes. As AI replaces this labor, explicit alignment controls weaken and systems drift from human preferences. Interdependent misalignment across institutions could become irreversible.
Goffman's framework reveals that LLM-based dialogue skips corrective rituals, entrainment, adjacency pair accountability, and co-presence cues that humans use to build trust and repair understanding. This ritual gap explains apparent fluency masking actual communicative failure.
Research identifies a distinct dialogue type where both parties modify their positions through exchange until compatible but not identical. Current AI systems collapse this into false agreement or AI-wins persuasion.
Once agents hold credentials, transact value, and interact with other agents, raw model capability stops being the limiting factor. The real bottleneck becomes whether agents can coordinate reliably, settle accounts, and leave auditable evidence of their actions.