I meant to write that the discrepancies between monthly, early reporting, and later, quarterly reporting, have been a feature for years. Maybe decades. Much of this is structural, survey reporting v statistical reporting, but it's there. An astute observer would generally discount monthly reports knowing thewy are regularly revised in the negative.

I suspect those who have a serious, fiduciary interest in using these reports do indeed rely on the later, quarterly, 'revised' reports, and spend a fair amount of time explaining to their clients the problems with monthly flash reports. It's the problems of restating several months' worth that is of more interest to me. This is an issue of trust.

Yes, fair enough.

"early estimates have been consistently too *optimistic* for many months."

Maybe for many years, in total.

You trust 'them' to operate ML systems better and more objectively than any other system?

Software isn't true, ethical, honest, or trustworthy. Those qualities are what programmers and their employees exhibit. Or don't.

I get concerned when the revisions exceed certain margins. Especially when they restate by 25-50%.

Actually, does anyone who needs accurate data believe the initial reporting for any month? Doubt it.

In 2000-2002 I managed security and infrastructure for a high school in Maine. This was a school with many very talented and diligent students. And some were learning to program in Turbo Pascal.

Someone either decided to, or did not notice the risks of, included the network libraries, something that was optional. I asked the publisher, and they confirmed, most school systems did not ask for that.

Sure enough, some students succeeded in writing a new GINA, intercepting attempted logins, and boom - they got a teacher's signon credentials to the NetWare server etc. And wasted no time rewriting some rows in the school grading and scheduling app. Much hilarity ensued.

It took while, but I was able to write a script to readily identify a non-student login, coloring the screen background bright red (it was NetWare, after all), and exposing the perp. The second part of this was a pair of cable scissors for the lab monitor or teacher in attendance. And the script dumped keystrokes to the equivalent of /dev/null for 10, later 15 seconds, trapping the perp at the screen. Some students power cycled, but most were a little stumped.

Instructions to the staff were simple. If a red screen appeared, take the scissors over. If a teacher or staff was not near that station, find the blue cable at the back and cut it. Took me about 30 seconds to re-terminate ethernet cables back than, and I had ago kit to do so, happily if we caught a perp. Most I passed in the hall, and they knew who I was, giving me some mean looks.

Solved in a few days. We left in the goodies, and I was tasked with giving a class on NetWare basics, oh fun. Which led to a much more devious intrusion, and a student being threatened with expulsion from they computer lab for the remainder... Their parents were outraged, this student needed those courses to get into the Ivy League school he was accepted at, early, (Legacy), and this was intolerable. My complaint was capturing data form backups to restore data that had been changed. And a brief discussion, which I led, pointing out that if such unethical if not illegal activity did not disqualify their son from entering that fine institution, well, my esteem for that institution was diminished somewhat. They curbed their child, and we got through the year. with no further detected breaches.

This isn't new, and this sort of behavior predates personal computers. But it is fun. They got to learn somehow.

"We ran a climate model to reconstruct the historical period, and then we ran it again but without the emissions of a specific carbon major,"

In other words, you took an educated, or at least informed, guess at the past, Then tried it again with some other assumptions.

Really. Guessing?

Sadly, you may think you're special.

Recent events prove you are not. And you are not like me, for sure.

Eliminate the Department of Education. Summer the necessary work elsewhere. The results argue for elimination.

FBI, DOJ, CIA, both the transgressions and results argue for cuts, 50-75%. If the case is made, ok.

Entitlement programs don't cut the benefits, they cut the true waste. Simplify regulations. I see the VA as an entitlement program.

Has anyone asked why the major military branches have individual major hospitals in the DC area?

Given recent revelations about days, I'm wondering WTH the Department of Labor is actually delivering for results.

But, bottom line, an unsparing review and assessment of our federal government.

Oh, repeal the 17th Amendment, and institute term limits in Congress.

I don't support tax increase for anybody. I super cutting the federal government by half. At a minimum.

Our federal government has been in a growth curve for almost 24 years, first post-9/11, then 2008-9, then 2019. Cut it.

As a side benefit, strip federal responsibility to the necessary. Keep them busy, dismantle the surveillance state, just do it before we cannot.

It may already be too late.

Ditto...

0) when it's a burden to report you activities and financials to current and prospective investors you should rethink your raison d'etre.

1) more information, not less, for investors.

2) if your management is chasing quarterly reports they are doing that wrong. Unless your strategy is to attract investors. If your strategy is to deliver long-term success, well, quarterlies don't matter as much.

Ditto in Maine and Acadian Canada. Even Newfoundland. And by now, certainly southern Labrador.

Hunger has not been a problem of supply for most of my adult life. Starting in the 60s. The US sent "surplus commodities" and much more overseas.

The real problem has been, during that time, war and socialism. And this continues. Climate change may indeed affect food production, but getting even a little bit of it where it is desperately needed is the real problem, has been, and still is.

ps - I still believe growing food for fuel is wrong, but that is a corollary to my initial point food is too often not where it is needed.

Well I'm encouraged that you do. Enough so that I, too, care enough to comment.

I'm not at all sure that 'most communities' pump sewage into rivers, most have at least minimal treatment plants. But if you, AC, have some documentation, please, share.