944 karma · joined October 10, 2006
(As an analogy, in theory there are tons of different ways to construct non-measurable sets, but in practice all examples come down to some quotient of a nonzero measure set and a measure zero set, like [0,1]/Q.)
Peter Thiel's analogy of this is "what's true but few people agree with you on" and pg's version is "the best ideas look like jokes / bad on first glance". There's a popular venn diagram that's almost analogous to the one in the article: the best startups are at the intersection of true and not obvious.
This article maps "obvious good ideas" to "provable in a formal system", which makes the Godel analogy work, but the Godel analogy seems like a worse analogy.
Analogies are supposed to put a new concept (startup ideas) in terms of more accessible concept (Godel statements?). The analogy target (Godel statements) was so obtuse that the author needed to spend pages explaining it.
Also, the Godel analogy is misleading. The canonical Godel sentence is self referential and also refers to the formal logic system -- a good startup doesn't need to be self referential or refer to the formal system. Besides generating Godel's incompleteness theorem, I don't think Godel statements are "interesting", unlike the most recent successful startups. (In fact, what's interesting are simple yet difficult theorems that were totally provable within Peano arithmetic to begin with, like Fermat's Last Theorem.)
A quick Google result shows that 592 total episodes have been filmed of the show. Under the null hypothesis that there was no cheating in any of the shows, you would still expect more likely than the chance that some show has a strong p-value of 0.3% for cheating to be 1-.997^592 = 83%.
As a jury, I don't think I would convict with just a p-value of 0.3% alone. You'd want something like a DNA test with a p-value expressed in 10^-6 or something to reduce the problem of multiple hypothesis testing / setup assumptions. Or you'd want to rely on other evidence like a confession or wiretap.
Therefore, this makes it less likely that there is a paradise lost of high restaurant margins we can return to, and more likely that some fundamental parameters that make the economic equilibrium of restaurant margins really low.
A guess for these parameters include:
- High competition: there are dozens if not hundreds of different restaurants I could walk to in 10 minutes (living in a US city).
- Low barriers to entry: a good minority of the population can learn how to operate an okay restaurant.
- A lot of owners get "passion utility" out of running a restaurant -- this means they are willing to be paid less in cash.
- Lastly but maybe most importantly, incredibly price sensitive customers. Just like the retailing industry, individuals are incredibly brutal about price. We can exhort day and night about the bad practices of Amazon, Walmart, boycott country X, but at the end of the day price wins. The author here is making a "boycott low margins" appeal which doesn't seem practical.
Restaurants will re-equilibriate. Likely a bunch will go out of business sadly. If social distancing requires there to be fewer customers, and customers are more cautious and eat out less, the bright side of the thin margins is that the restaurants will just pass this cost onto the customers since the ones that don't will be out of business.
Since all restaurants will automatically cost more resources (square foot per customer, overhead) to operate, they will have a common force uniting their upward push in prices too.
[1] https://www.forbes.com/sites/sageworks/2014/06/22/us-restaur...
Also, current consensus, of course among mainstream economists, is that countries that more believe the Austrian school, like Germany, has caused unneeded pain on themselves/their neighbors by advocating of austerity versus stimulus during downturns. If you want the mainstream steelman against the Austrian school, you can search Paul Krugman's take on them.
However, if you take the theme of Levine's articles too literally/thematically, you'll think that finance is filled with quirky rules that produces comically inefficient outcomes most of the time. You'll think that there are easy system changes that any layperson can see that would make the system run much better. You might even think that finance is clear swamp ready to be drained -- this plays into the confirmation bias of readers who come in already with the thesis that "the system is totally broken".
In fact, against this theme, 90% of the time the story is incredibly dry and uninteresting because the system works exactly as intended. 90%+ of the time, stock A + B bundled is worth exactly the sum of it's parts. 90%+ of the time, the regulations do protect investors, are cheap to check off, and causes no weird behavior. 90%+ of IPOs are textbook win-wins and not Adam Neumann playing Softbank or a sell-side banker eating grandma.
To be constructive though, what's the solution? Start with skimming over the most classical textbooks (Mankiw or the recommended non-heterodox textbooks from the relevant classes by your local colleges). This is the classical theory of both how things are supposed to work, and how 90% the value generation actually occurs in 90% of the cases.
If the bar is actually real gold, the nature of the damage is different. It's less about undermining gold than it is about undermining the brand of the stamp.
I am surprised by the amount of comments here that seem like they are more or less arguing against enforcement of DUI laws that will save lives.
Looks like we're at nearly an all-time low for the last 25 years as well, no?
To be fair, I think there is some argument that some of the larger tech giants would benefit consumer welfare by being broken up. Facebook probably less so since people already have choice.
Warren does herself no favors by using $25 billion as the threshold of a monopoly. What a real monopoly is is when there is no consumer choice at all, leading to large corporate profits that hurt consumers at the pricing level.
By not using consumer choice or harm as the standard (which is what the court uses), Warren comes across (and perhaps actually intends) to be against large companies, regardless of welfare.
And by the modern style, I meant starting out with analysis, defining the axioms of what measures are, demonstrating the existence of nonmeasurable sets with the axiom of choice, etc. IIRC, the course did rely on Borel algebras for its buildup, but did not openly buildup from sigma algebra machinery.
Having 1) studied "naive statistics" when I was young and 2) learning the full-on measure-theory version a la Bourbaki, and now 3) using applied statistics every day, I can say the naive version is the most useful by far. Yes, there is hand waving, but the hand waving works, and you really don't get into that much trouble unless you try to be pathological.
I admit, it would be great to have a version where the naive intuition (many of which are motivated by empirics, physics, and real world situations) and the theoretical definition matched better. Even still, I don't think the formal treatment is appropriate as an introduction -- it's an unnecessary hazing on learning minds when they can get most of the value with a bit of hand waving.
For example, the anchoring bias is actually a useful heuristic in many situations. If you ask someone about the length of the Nile and a third person guessed 5000 miles, you probably would guess closer to 5000 miles to integrate the third person's information, which is usually somewhat informative.
The bias comes when an outside marketer manipulates you by saying "how much should the iPad cost? $2000? No! $1000? No! It's only $500!".
So basically this lawsuit has already been dismissed by a lower court, and but Lufthansa is appealing. The entire dynamics here seem like Lufthansa is unlikely to win and even if it does the amount it is claiming ($3K?) is way lower than legal costs, but the publicity of the lawsuit is the entire point they're making. A version of SLAPP, if you will.
Roughly, it goes as such:
1) the probability of N being a perfect square is proportional to 1/sqrt(N).
2) For any N_0 arbitrarily high, if you integrate from N_0 to infinity the expression (1/sqrt(N) dN), you get infinity.
3) The expression in 2) is the "Feynman equivalent" of the expected number of square numbers above N_0.
So Feynman's nonproof actually turns out to be true, despite it not being a proof in this case as well.
First, I asked, and they denied it. Then I literally had to pull out this law and they finally granted it. But they did everything they could to obstruct me. Literally they refused to scan it and wanted me to show up in person. Then while writing it, the doctor himself said "sad to see you don't support our business". (For people living in Boston, the place was Custom Eyes in Central Square.)
I was surprised by the divergence between the law and how pissy the doctor got when I tried to exercise my rights.
For example, Princeton a nonprofit can totally charge tuition and use it as a base of funding, and not pay taxes on that. They can even sell their students food in cafeteria and not pay income tax. [1] But they can't start a food franchise and not pay tax.
This software seems clearly within their registration statement. Now if they started selling software consulting as their major source of funding...
In any case, I agree checking with your accountant is the best.
A Google search shows Google has about 40K software engineers, while the USA has about 18M software engineers. This means Google hires about 0.2% (i.e. 0.002) of the software labor force.
This means that Google only captures 0.2% of the benefit from increasing the labor pool. 99.8% of the dollar benefit goes to other people -- some of who are competitors even!
Further, Oracle hires a lot more software engineers, but Oracle is notorious for its profit motive and bad image while Google tries (in name at least) to be altruistic. Is Oracle spending as much as Google on STEM? How about IBM, SAP, etc? If the math works out for Google, it should doubly work out for companies employing more software engineers.
This is a theory that seems like it can be true (and make you seem smart at cocktail parties), but the math doesn't seem to add up.
Especially useful would be to see a burn-down chart of say, new smokers and new vapers of the same age range, and seeing cumulative lung disease and mortality, and seeing how the progress of that differs (or is the same).
I wonder who's paying the dinner bills after today's hearing.
Some people mentioned that Wag should have offered more and demanded less -- absolutely!
But think about if Wag even said "sorry this happened, that's all we can say for your terrible loss" even without a compensation offer and without an insulting NDA. It probably wouldn't have pushed the owner to post this and for it to end up on HN -- and now the world knows.
I honestly believe Wag and Rover and other large companies that offer pet walking services should put up a very high liability bond for pet injury -- on the order of $10K or more. That way they have skin in the game to vet their walkers better. Most owners value their dogs above that, and that incentive is the only way to get these corporations to listen.
Maybe this happens a lot more often than documented. In this case, it can throw the field of mtDNA testing and implications into question.
Conversely, maybe it happens extremely rarely and only in pathological cases (in the mathematics/statistics sense, if not the medical sense). In this case, it's just a medical curiosity and only important insofar as a doctor treating a patient with a rare disease may want to consider the remote possibility of this.
In the article, someone dressed as a hot dog doing marketing is misemployed because it's not "contributing to human welfare and flourishing".
If that's the case, then the century-long history of employment in post-industrial-revolution society is a story of misemployment. Is tightening a nut repeatedly 40 hours a week under the job guarantee of a 1950s GM true "flourishing"? And is the 1850s Irish farmer racing to harvest blighted potatoes misemployed as well? (Perhaps not, since the survival state creates an certain "authenticity" versus the "banality" of guaranteed life?)
To me, crapware is:
- Order of magnitude bad engineering by industry standards (esp if it is not customer centric).
- So bad that if it could be removed, a majority of people who know how would remove it immediately.
- So bad that if [FANG / anyone competent at software] designed an alternative, most people would switch to it.
Crapware for me is not identically equivalent to mandatoryware. I get the GNU-like hate for mandatoryware from some people, but it's useful to have a distinction between (potentially subpar) mandatoryware like IE in Windows XP, and the absolute rubbish that was Verizon Music Store on my 2005 flip phone.
This brings up the point: if the cells can indeed divide indefinitely, which even stem cells cannot, then aren't these effectively cancer cells / made cancerified? Perhaps this was done on purpose independently to make the cell culture easier to work with, but that should have been mentioned.
For example, what does one learn from a successful CEO who is dominant, unempathetic, surrounded by yes men, and fiddles with their phone?
Perhaps I'm not reading this correctly, and you mean look at these traits and note how (perhaps counterintuitively) effective or ineffective they are?
The closest cheap thing is probably the few seconds you get on a trampoline.
Given that they built just one testing rig, and they use information like "commonly known object sizes" for depth, I wonder if they are using any transfer learning at all? I imagine whoever is bringing that rig around can't look at all objects, and transferring the knowledge of "common object sizes" and even what an object is can be hugely valuable for such a project.
I bet this is also a sneak peak into how powerful self-driving cars can be at Waymo. Not only do such cars have access to human data (sights, sounds), they can have way more sensors which can combine in a neural network to predict things way better than humans. For example, a NN could infer the chance a pedestrian might veer from the sidewalk into the streets by analyzing:
- The psychological state of a person using a high res camera on the face.
- Using chromographs to see if the pedestrian is drunk.
- Infrared cameras to see the person's temperature / if she's excited, etc.
- Limb modeling and likely walking paths.