sirsar
> Crop yields increase 3-5×
> Farmers go from $600/acre to $14,000/acre revenue
5×$600 is $3000. Where did the extra 4.7x come from? The new-to-the-world info looks more like "making stuff up on the fly".
> Fatal atherosclerotic events were more common in the control group (70 v. 48; p<0.05). However, total mortality was similar in the two groups: 178 controls v. 174 experimentals, demonstrating an excess of non-atherosclerotic deaths in the experimental group. This was accounted for by a greater incidence of fatal carcinomas in the experimental group [the vegetable oils group].
Read as (Apple silicon)-based.
It seems to me there's value in posting the details before the results are in:
- Helps prevent publication bias, by encouraging posting negative results
- Elicits comments from knowledgeable people who may be able to help
- Discourages doctoring the story after the results are in
> Even fewer can make a safe vaccine.
I don't know, the article, whitepaper, and comments seemed to make a pretty convincing case for the nasal vaccine attempt here to be not very dangerous - probably useless at worst. Do you have a more substantive counterargument?
> Consider this a pre-registration. I intend to share my test results here.
How about, instead of totally dismissing this - in classic HN fashion - with "this can never work" or "this is magical thinking", we let the guy share his results?
If it works, that will be a massively important result. If not, it won't be surprising, but _at least he will have tried_. I for one think the expected value leans overwhelmingly in favor of at least one person actually trying.
But I live in Cambridge, MA and I can assure you Somerville would remain one of the densest cities in the union if every last building therein was lopped down to 3 stories by a giant lawnmower.
The key is to realize the cars themselves killed density.
Parking lots, parking spaces, and extra lanes all conspire to push humans and human spaces further apart. This then makes walking less feasible, cars more required, and more space required to accommodate those cars in a feedback loop.
Taller buildings are not required for density, far from it. Look at Somerville, MA, where just about nothing is higher than 3 stories, yet they fit almost 20,000 people into a square mile -- and they like it. What they don't fit is 20,000 parking spaces, and that makes all the difference.
1. The appeal of the city is drastically reduced by coronavirus + remote work 2. Boston cost-of-living is really expensive
Put them together and moving to New Hampshire or Vermont starts to seem really attractive. Just compare Zillow for Manchester, NH vs, e.g., Somerville, MA.
Combine with 3. Most MA residents live in Greater Boston, and it's easy to see MA at #5.
If I'm searching for a murderer in a town of 1000, it takes about 10 independent bits of evidence to get the right one. And when I charge someone, I must already have the vast majority of that evidence. To say "oh well we don't know that it wasn't Mr. or Mrs. Doe, let's bring them in" is itself a breach of the Does' rights. I'm ignoring 9 of the 10 bits of evidence!
Using a low-accuracy facial recognition system and a low-accountability lineup procedure to elevate some random man who did nothing wrong from presumed-innocent to 1-in-6 to prime suspect, without having the necessary amount of evidence, is committing the exact same error and is nearly as egregious as pulling a random civilian out of a hat and charging them.
If you're going to report other ratios as percentages, could you please be consistent? I initially erroneously read this as 0.01% deaths, which would be an absolutely enormous update, but 1% isn't surprising at all.
There is some price per mask at which the N95 respirator manufacturers that currently exist in the US would be willing to hire more people, buy more supplies, etc and start running 24/7. The only reason they aren't already is because nobody has agreed to pay for it.
I held the same perspective until I worked for a manufacturing company. Many of that company's products, and indeed many products throughout the world, _would not exist_ without market segmentation. Selling the same hardware with different software often makes the company more money than only having one model. In some cases, that extra money is the difference between the product, or even the company, existing or not existing.
R&D and manufacturing are hard and expensive.
Slate Star Codex [1] has a good overview, but doesn't provide a solution (instead only providing some good evidence that rejects some hypotheses).
As far as I can tell, the single biggest driver of Cost Disease is labor, but not salary: it takes more people to deliver the same services. This in turn is driven by market failure:
- The signaling component of higher education is zero-sum, and infinite loans and foreign capital mean not enough few consumers make college decisions based on cost
- US states have shown insufficient teeth in enforcing caps on construction budgets; overruns are expected meaning the bidding process cannot actually cause cost savings. And the days of private light rail companies competing against each other are long, long gone.
- Primary education's cost growth has been concentrated in areas that are not teaching - administration and bureaucracy. There is, of course, no meaningful cost competition at all in primary education.
All of these carry the caveat that we used to be able to get these things for cheaper, and e.g. primary education was never traded on a free market. So what happened? Was it inevitable, and merely took time?
If anyone has other ideas, please let me know. I would love to vote against cost disease and call any politician who will listen, but I don't fully understand the problem, let alone have a solution to offer.
[1] https://slatestarcodex.com/2017/02/09/considerations-on-cost...
- You can create public pre-signed URLs with an unlimited expiration date using SigV2. I doubt SigV2 will stop working anytime soon.
- Or, if you want the link to provide access to anyone, why not make the object public, but put a UUID in the key name to prevent enumeration?
I suspect the answer is something having to do with the overlap between the politically powerful and those with a great exposure to those risks and rewards.
> Open to having my views changed on this.
If you agree that a world with more cyclists would be better (for the environment, for safety, maybe for public health), how do you suggest we get there? Encouraging change towards a better world often requires swimming upstream, shouldering some stress and discomfort and pressure, in order to exert pressure back on the system one wants to change.
Nobody will build safer bicycle infrastructure if nobody rides bicycles.
Even without doing any math, or proofs, or having any algorithmic background, it shouldn't take "fumbling about on some recursive goose chase" to write the function from first principles:
>>> def games(players):
... if players <= 1: return 0
... if players == 2: return 1
... return 1 + 2 * games(players // 2)
...
>>> games(8)
7
[0] To me, trick questions involve some hard-to-come-by "aha" moment, like the XOR trick to swap two variables without a temp. This one doesn't seem to have that.
That on top of its health benefits, relative safety to those hit by bikes, its traffic efficiency, parking space efficiency, benefit to retail...
You bring up taxpayers. It is far, far cheaper to subsidize a bike lane than a new subway tunnel. And the bike lane is open more hours of the year, with more entrances and exits, more interchanges to other lines, a shorter distance point-to-point, with drastically lower maintenance costs. Taxpayers should be rioting over the costs spent on trains and traffic relative to bikes.
Unfortunately for this theory, gauge blocks will remain wrung together even in a vacuum!