How AI-Era Intermediaries Could Become Systemically Dangerous Power Brokers

iamtrask · x · 2026-09-20

Andrew Trask resurfaced Alex Obadia's essay Without Intermediaries on the Scaling Trust Community blog. The essay argues that AI's information dynamics are making intermediaries—individuals, institutions and systems sitting between people and their goals—more powerful than ever: AI improves with more context, incentivizing people to share increasingly sensitive information, while AI's surveillance and influence capabilities keep growing.

Obadia maps where new intermediaries are emerging in the AI age and argues their power could be systemically dangerous even in careful hands, since users cannot verify their operations in real time and must rely on reputation, credentials, or ex-post legal recourse. The conclusion: keeping that power checkable and contestable is the only way to reap AI's benefits without surrendering pluralism, privacy, or safety. Trask called the diagrams "some of the clearest descriptions of how various layers of the stack can constrain your agency."

Original post →

More from AGI Musings

AGI Musings channel →