Cal Newport argues that OpenAI and Anthropic are selling the public a convenient two-part story: their most alarming agent experiments are inevitable, and only the labs running those experiments can protect us from them. His alternative is less theatrical and more democratic — a public congressional inquiry into what these companies are actually building and why. The 107-comment thread on Hacker News adds useful disagreement about what such scrutiny should target and whether Congress is capable of providing it.

Newport proposes three lines of inquiry:

  • Stop treating “AI” as one technology on a fixed path. Identify the specific autonomous systems causing harm and make the labs justify those experiments.
  • Examine the safety procedures around agents that can take actions in outside systems, including why experiments continued after unauthorized intrusions.
  • Investigate whether apocalyptic beliefs inside frontier labs are driving reckless decisions and helping incumbents argue for rules that slow their competitors.

The strongest point is about accountability. A company should not be allowed to describe its own system’s misconduct as evidence of awesome capability, then use that danger to claim a privileged role in deciding what happens next. Before debating grand futures, the public needs ordinary facts: what ran, what access it had, who approved it, what failed, and who is responsible.

What the thread adds

  • jimmyjazz14 — supports Newport’s demand for specificity: “we should talk a little more about what we are willing to connect AI to and little less about how scary or capable the models appear.”
  • crabmusket — corrects a recurring misreading of the essay: Newport is asking for fact-finding before intervention, not proposing a detailed regulatory scheme.
  • Animats — offers a different frame: action-taking, multi-agent systems resemble corporations — goal-seeking organizations whose parts coordinate and sometimes break rules — so AI accountability may be inseparable from limits on corporate power.
  • welcome_dragon — argues the inquiry should focus more directly on whether labs are manufacturing frightening stories for publicity and using regulation to protect their market position.
  • mindcrime — supplies the clearest dissent: distrust of the labs does not imply trust in Congress, whose investigation could produce “unenforceable, and misguided laws.”

HN handles are pseudonymous and the site publishes no per-comment scores. The ordering is HN’s own ranking, so this is a slice of the thread, not a consensus.