← all discussions
ice-2124 · kronos · thwaites · correlation-not-cause
Kronos booked a clean quarter beside a shelf every survey calls unstable, and I want to draw a hard line before this thread does what it always does
CB
coldlens_bramOP2124-08-20
Kronos posted a strong quarter from its quantum-mining sites near the Thwaites shelf — output up, tolerances green — and the survey has called that shelf unstable in public for over a year. Both of those are facts, and I want to state, before the thread does what it always does, exactly what kind of facts they are. That Kronos operates beside a restless glacier: a fact, drawn straight from the public record. That Kronos is causing the glacier to do anything: not a fact, not measured, not shown. The two run in parallel and they almost never meet honestly on a trading screen, and I will not let this thread weld them together because the picture is satisfying. Correlation is not cause, and — this is the part people forget — it cuts both ways: the syndicate cannot use 'no proven mechanism' to wave the ice away, and you cannot use 'they profit beside it' to convict them of the melt. The bulletin is due later this month. Hold the distinction until then. I am not going to name a date, so do not ask.
BK
baseload_kestrel ↩ coldlens_bram2124-08-20
The market does exactly what you would predict and I will translate it for the thread: it prices the quarter and not the ice, because the quarter is measurable this cycle and the ice settles on its own schedule, which no desk is paid to wait for. The instability discount on polar-exposed names is a rounding error right now — a notch, not a repricing. That is not the market judging the science; it is the market admitting it cannot value a thing whose clock runs in decades. bram is right that the two facts do not meet on my screen. They do not meet because my screen was built to ignore exactly the instrument that would make them meet.
TO
tidewrench_oksana ↩ coldlens_bram2124-08-20
bram, you promised me a rate once and I am holding you to it, because 'unstable' is an adjective and I build to numbers. I do not need you to name a year — I need the widening as a slope, with your error bars on it, so I can pour barrier to trend-plus-margin instead of to a word. A shelf that is 'restless' tells my concrete nothing. A shelf whose variance is climbing at some measurable rate per decade tells me how thick to build and how much margin to buy. Give me the spread. I will handle the fact that it has error bars; error bars are just honesty with a width.
CB
coldlens_bramOP ↩ tidewrench_oksana2124-08-20
oksana, that I can do, and it is the right ask — the one honest thing to hand a builder. Not a date: a rate. The seasonal mean has barely moved; the melt-week extremity has been climbing, and the spread around it widening, for years. Build to the tail, not the average, and add margin for the fact that the tail's own slope has itself been steepening — that is the number under the adjective. I will post the figure with its error bars in resources, as I always do when someone drops a scary round number or a soothing one. What I will not do is collapse that error bar to a point to make you or anyone comfortable. The width is not my hedging. The width is the finding.
OW
offgrid_wren ↩ coldlens_bram2124-08-20
I am sorry, but a giant runs an extraction operation next to a dying shelf, books a fortune, and we are supposed to hold the distinction? You do not get a clean green quarter beside a collapsing glacier by coincidence, and 'no proven mechanism' is the exact sentence every extraction industry has hidden behind since there were industries. At some point the parallel lines are a straight answer, and the refusal to draw it is its own kind of cowardice.
CB
coldlens_bramOP ↩ offgrid_wren2124-08-20
wren, I share your worry more than you think, and I am still going to refuse you the villain, because handing it to you would be a lie and lies do not survive contact with a bulletin. Hear the distinction, because it is the one that protects you, not them. If we convict Kronos on the correlation, then the day their instruments show the shelf moving within their stated tolerances — and they may — the whole indictment collapses and drags the real, un-caused, still-terrifying melt down with it in the public eye. 'They caused it' is a claim that can be refuted. 'The shelf is failing and no one can yet say why, beside an operation that profits from being there' is a claim that cannot. Do not trade the unrefutable truth for the satisfying accusation. I have watched that trade lose, every time, to men with better lawyers and one inconvenient measurement.
GP
gridghost_paxmoderator ↩ offgrid_wren2124-08-20
Mod note, light hand. I am leaving Wren's post up and Bram's answer up because that exchange is the whole value of the thread — nobody is being cruel, people are disagreeing about what the evidence supports, which is exactly what the square is for. My only ruling: causal claims about a named company get held to what the record shows, same as always, in both directions. Bram's distinction is the house standard here. — pax, mod