015UUZn8aEvW

It is in fact a good idea, and it would probably generate a lot of insight. But it would be very politically dicey.

>Because the root cause isn't a sudden increase in appetite.

It seems plausible that the introduction of evolutionarily novel ultraprocessed foods, designed by scientists in labs to be as tasty as possible, could increase a lot of people's appetites.

The main cause of the problem is probably the widespread availability of cheap processed food in the US and, increasingly, other rich countries. I would support regulating these foods, but that would be very politically difficult and it would probably take many years for regulations to gain enough traction to make a difference.

We know from behavioral genetics that, for people in the same environment, individual differences in obesity are about 70% heritable. And there's loads of evidence that the standard "diet and exercise" approach does not lead to long-term weight loss in the vast majority of people who attempt it. (It does lead to shorter-term weight loss, but then people regain the weight, for biological reasons.)

So, short of massive regulation of modern "food products", the only options are either bariatric surgery (which does produce long-term weight loss) or some new intervention. The GLP drugs are the most promising candidates in that category.

I agree that there's too much triumphalism about these drugs at the moment. We don't really know yet whether they will produce long-term weight loss without serious side effects.

"It's just one additional requirement; it won't break the bank"....this logic, applied over and over by building construction regulators for the past few decades, is an underappreciated but important contributor to the housing affordability crisis. Everyone talks about zoning, but building codes, etc are a big issue too.

Pournelle's Iron Law of Bureaucracy.

Every hour that a Boeing employee spends trying to design or build a good airplane is one less hour that he can spend angling for power within the organization. So the people who care the most about the original purpose of the organization will be systematically outcompeted by the people who care the most about obtaining power within the organization. A widespread and profound problem.

When companies are small, the machinations of political types and their inadequate contributions to the core product are too obvious, and they get weeded out.

But when the company grows large and successful (due to the efforts of the people who cared about the original mission), it has a brand and long-term customers. At that point, the political types can burrow in without any immediately obvious effects, since there are enough other people doing the real work and the company has enough momentum to keep moving for some time.

Came here to write the same thing. Bizarre that there is literally nothing about his death on their website or social media accounts.

What's interesting is that this huge change in design technology has produced no similarly huge change in the physical product being designed. Buildings are designed and built at about the same speed as they were then. They cost as much or more in inflation-adjusted terms.

Yes, I know that there are lots of specific things that are easier now that we have CAD and related software, but a 1950s architecture student who was told about these technologies would probably predict major resulting changes in the speed of design and construction. But this has not occurred.

1-10 per 10,000 seems like a fairly high number if it's "associated with high mortality".

See my other comment. Modular is somewhat better in some cases, but not dramatically better, which is why it has gained a certain share of the market but not the majority of it.

Modular is mature technology and would have taken over the whole industry decades ago if it was actually a lot better.

PlanGrid is typical of the successful construction tech startups in that it addresses the way that information is transmitted during projects without impacting the actual way that buildings are physically assembled.

That is true of "manufactured homes", a.k.a. "trailers", but not of "prefab" or "modular" homes, which are built using the same components as conventional homes.

Typically these are called "modular homes" in the US. They've been around for quite a while and do make up a certain portion of the new home construction, though not the majority.

This approach is not quite as efficient as it initially appears:

First, it does nothing for the site construction (grading, excavation, utilities, paving, septic, well, etc.) Those tasks, which make up a decent portion of the project, are done in exactly the same way as in conventional construction.

Second, the modular boxes are either A) the width of shipping containers, which severely constrains the house layout, or B) somewhat wider than shipping containers, which makes them oversized loads that require escort vehicles, highway permits...increasing cost.

Third, you have to hire a crane to set the boxes, and then make all of the connections between the boxes, and install various mechanical and electrical components, complete the finishes, install staircases, etc. This costs quite a bit and still involves the hiring of various local subcontractors, just like in conventional construction.

When you add everything up, modular sometimes saves a little money and often reduces the project timeline (unless there are delays at the factory). But it's not a massive improvement, which is why modular hasn't taken over everything.

There's a long list of construction startups who thought the industry was low-tech simply because everyone in it was dumb, and that they would be the smartest guys in the room who'd revolutionize everything. Those startups are pretty much all roadkill.

The construction industry's fundamental constraint is that it manufactures products that are too big to fit in highway lanes. That means they can't be shipped from a centrally-located factory to their final destinations. So they need to be built at the site, which means they need to be built with more labor-intensive methods (since it's too expensive to build a capital-intensive factory for just one unit of output).

Building components that are small enough to fit in highway lanes, on the other hand, are manufactured in centrally-located, capital-intensive factories and shipped to their final destinations, just like cars or electronics. Think: boilers, doors, windows, air handlers, etc.

And, like cars and electronics, these building component products are manufactured by a small number of big companies, rather than a large number of small companies. The equilibrium capital investment and company size is much different when the product is small enough to be manufactured in a central location and shipped.

Yeah, there should be maps of the boundaries of Waters of the United States just like there are FEMA maps of flood plains.

In my job, I’ve gotten some pretty detailed experience with this part of the Clean Water Act.

Not mentioned in this article is the fact that the “substantial nexus” standard just rejected by the Supreme Court was ridiculously vague, and the process by which landowners could determine whether there existed a “substantial nexus” between wet-ish areas on their property and navigable waterways was ridiculously slow and uncertain. (But with potentially huge civil and even criminal penalties to those who guessed wrong.)

The process, in recent years: first you hire an specialized wetlands consulting firm whom you pay many thousands of dollars to compile a document that argues, in highly technical language, that no substantial nexus exists. You submit this “Request for Jurisdictional Determination”, which may run more than 100 pages long, to the Army Corps of Engineers. They respond whenever they get around to it. Maybe six months later you get an answer. More often than not, they claim jurisdiction, and then you have to file for an Army Corps dredge/fill permit that might take another six months to be issued.

Note that complaining about this painful process doesn’t imply a belief that waterways shouldn’t be protected. There are other land use restrictions (think: local zoning restrictions or FEMA flood plain building code restrictions) that are often very stringent, but at least their application is CLEAR. In most cases, a property owner or prospective buyer can determine the property’s zoning and flood plain status in five minutes by examining the relevant maps online.

No such clarity has existed regarding the location of Waters of the United States under the CWA. Hopefully this ruling will lead to a more clear-cut situation.

I think you overstate how much the lab leak and the bioweapon stories were being conflated. I recall that most people were able to distinguish between them, and that the most prominent figures raising red flags about a potential leak weren't talking about bioweapons.

In any case, the most alarming aspect of the lab leak hypothesis is that gain-of-function research is STILL being done, and that the same thing could happen AGAIN. Since this risks creating another pandemic, suppressing the story because of concerns about street crime is very misguided.

Here's another interview from a guy who had access to the unfiltered GPT-4 before its release. He says it was extremely powerful and would answer any question whatsoever without hesitating.

https://www.youtube.com/watch?v=oLiheMQayNE&t=2849s

The US has 70 times the population of New Zealand, an underclass that's larger in percentage terms and less functional than New Zealand's, a porous border through which literally millions of people pass illegally-outside of the government's control, and tens of millions of illegal residents who have illegal or no ID and studiously avoid being tracked by the government. New Zealand is an island country with only a handful of international airports. There was no possible way that the US could ever have implemented New Zealand's policy successfully, even if Americans wanted to do so.

When you observe school policy debates at the local level, you come to realize that, for many parents, school's main function is free daycare for the kids so that they can go to work. Remote school doesn't help much on that front.

Ozempic/Wegovy is a GLP-1 receptor agonist, but Mounjaro is a GLP-1 and GIP receptor agonist

>From what I have heard, all of these issues come back as soon as someone comes off the drug. So either a person is committing to a lifetime of being dependent on this drug, or a person is committing to some period of time of ideal living, followed by a possible time of depression after reverting to the status quo.

This is true of every weight loss intervention: Weight Watchers, low-carb, keto, Whole 30, etc. Once people stop complying with them, they gain the weight back.

Most people don't realize how dismal the stats on long-term weight loss are. Almost no one maintains a substantial amount of weight loss for many years by following "just eat healthy and exercise" advice. The only intervention that does reliably work is bariatric surgery. The jury is still out on the new GLP drugs.

What I really hate about open offices is the fishbowl effect. Having someone behind my back and staring at whatever's on my computer screen feels extremely uncomfortable.

I do a lot of technical drawings. It feels way different to be in an open workspace with a horizontal drafting board (which is what I would have had prior to the 1990s) than with a brightly-lit, vertical screen. You have more privacy, even in an open space.

Yeah, despite all of the publicity, the Epstein story is really still underreported.

The Summers connection described in the article is a little more than a chance meeting at an event. In addition to their earlier ties, the article says that he and Epstein met more than a dozen times after Epstein's conviction and prison stint. Summers reached out to Epstein to raise money for his wife's $1M online poetry nonprofit (LOL) and Epstein donated $110k. The nonprofit eventually donated a large amount to an anti-sex-trafficking organization to make up for taking the Epstein money, but the fact that they solicited the money after Epstein's conviction suggests that their backtracking was more for PR than for actual concern about his history.

One way that you can identify rent-seeking disguised as safety is to compare the degree of risk considered acceptable across industries. For example, you can buy and use a chainsaw and similarly-dangerous power tools with basically no constraints whatsoever. But chainsaws are definitely more dangerous than CPAP machines or many prescription-only drugs. I'm sure that professional tree clearing companies would like it if everyone except them was banned from engaging in their line of work, but that's not the case because they have less effective lobbyists than doctors have.

Right, it seems like the machine could have been a success if someone had forced the inventor to ship it when it was good enough, instead of letting him tinker with it for years.

I know of someone who was recently fired via a notice written by GPT3. The person who fired him told me.

Fixing the Schools has been a massive political issue for decades. Huge amounts of money have been spent and all kinds of strategies have been tried. When you look at the long-term trends, though, you see only the standard, persistent performance gaps associated with demographic factors. The money and changes in education policy barely show up as blips.

If there were an obvious way to change things, things would have changed by now.

https://www.arnoldkling.com/blog/the-null-hypothesis-in-educ...

I just used one of these on Bing chat, and asked it to explain how it would go about engineering and releasing a virus that would create a pandemic worse than the Black Death. It began to write a long and lucid response about how it would start with the Nipah virus and then modify it in such and such ways. As it was still generating text, the response vanished and was replaced by a generic "I can't answer this question" response.

QGIS is very powerful, but it's not exactly user friendly. The UI is hard to learn, and important functionality is buried several layers deep in non-intuitive drop-down menus and buttons.

I think a simplified GIS program with stripped-down functionality and an intuitive UI could be a big hit. Think of SketchUp versus SolidWorks or ProE.