[Aeyxen]

The thread keeps circling around the politics, but almost nobody has dug into what actually goes on in the NWS tornado warning pipeline.

It's worth being specific: the National Weather Service operates some of the most robust automation and radar ingest pipelines on Earth, but the final go/no-go warning call is almost always human—often a single overnight forecaster on a console, monitoring a swath of counties. Automation (e.g., Warn-on-Forecast guidance) can surface threats, but the NWS intentionally doesn't have an 'auto-warn' button for tornadoes, because of the asymmetry of false positives (blow credibility, cost lives in the long run).

Budget cuts reduce redundancy and experience in those overnight shifts. When you have only one person monitoring instead of a team of two or three, you get decision fatigue and coverage holes, especially during clustered, multi-cell outbreaks. We've seen near-misses in the past, and every pro-meteorologist I know says they're playing defense against process errors, not just technology failures.

Before we point fingers or blame 'technology/automation' shortfalls, let's quantify the concrete bottleneck: skilled human decision-makers are the limiting reagent; machine learning warning aids are still years away from majority trust.

[johnnyanmac]

>Before we point fingers or blame 'technology/automation' shortfalls, let's quantify the concrete bottleneck: skilled human decision-makers are the limiting reagent

All the automation in the world with useless without a human guide to either transform production into a useful product, or useful knowledge to heed. That's why this act of trying to remove human labor is asinine. Even skilled human can't always get the right readings, so expecting a robot to do it all at this stage is just selling snake oil.

[Aeyxen]

Actually, I'll go a step further - in the long run, we probably won't need human forecasters at all.

The current human-in-the-loop model exists largely because our technology hasn't been good enough yet, not because there's something inherently special about human judgment in this context. Weather prediction is fundamentally a pattern recognition problem. Pattern analysis at scale is exactly what computers do better than us.

Perhaps someone could apply to YC with this idea. There is one YC startup doing this already: https://www.ycombinator.com/companies/atmo

[johnnyanmac]

Chaos theory and brownian motion make it a herculesn for anything to predict the weather more than a few days out. There's too many micro factors to track leading to weather that constantly shifts. And The data costs to attempt to try to do so is well past even the most well compensated meteorologist.

I'm not too worried about the human factor being replaced as a whole here. Even with AI, someone needs to interpret the output and make sure the the prediction models actually work.

[Aeyxen]

Yeah, true that.

[siilats]

yeah why cannot that guy sit in california or new york in a normal time zone? not like there are tornadoes in every state, its so silly to keep a person at night in an office when weather is good

[fmorel]

There's only a 6 hour different between the East coast and Hawaii. You can't entirely avoid a night shift, so you might as well have them all work from the same location.