papeda

I always rep Wikipedia's "random page" function [1]. Random page is especially useful if you are cooped up at home and can't access normal sources of randomness like close conversation with other people, the rustle of leaves in trees, sunlight filtering through passing clouds, etc.

[1] https://en.wikipedia.org/wiki/Special:Random

This article focuses on a delivery network powered by a large pool of cheap, unskilled labor. The role of poverty levels in India is very relevant to this topic. Your comment is also the only one that brings up caste.

+1. A slow, naturalistic movie that's also conventionally entertaining. Plus, as a timely feature, one of the protagonists spends a bunch of time standing on crowded trains, sitting in a crowded office, and eating lunch in a crowded canteen.

Thanks, this is a clearer expression of the argument I wanted to make.

I agree with your statement. The comment I was responding to does not:

> If anything this is evidence of Google’s search engine monopoly and Google built the selection feature into Chrome it in furtherance of that monopoly power

I see. I'm interpreting the DoJ's choice as a claim that Google is the most harmful monopolist in the US, since they've chosen to go after Google (and not anybody else). That's why I wrote my original comment. However, that's not actually stated in the article.

I think my interpretation is reasonable, but I get why it's whataboutism in a discussion of the topic "is big tech monopolistic?".

We seem to disagree on what the original statement is.

When I talk to my peers, the original statement is some version of: "big tech is the most harmful monopoly in the USA today". Saying "what about monopolistic industry x?" is directly engaging with that statement.

It's possible that I'm conflating people's decisions to prioritize talking about tech companies with their actual priorities for which industries require government intervention.

I don't see this as "good is the enemy of perfect". I see federal contracting as a worse force for waste and bad outcomes than tech companies, so I suggest that targeting that industry would better improve people's lives.

Not every suggestion that we've misidentified a big problem is whataboutism. Especially when elected officials (in theory) look to their constituents to determine which problems to address/industries to threaten.

> If this hypothesis were correct, we would have expected to see lawsuits against those other monopolies during, say, the Obama administration. But we didn't.

This presupposes that the DoJ has always been a political weapon.

> Healthcare has to do business with Google. Defense has to do business with Google.

Can you elaborate on this? Neither industry seems to rely on ads or SEO or app stores.

Hm. If I understand the cellophane paradox, it says that the "substitutes exist" does not imply "there is no monopoly".

That seems different from your argument, which I understand to be "only a monopolist would make it easy for consumers to choose a competitor". But again, the conclusion that Google is a monopolist seems to be baked into this logic -- would you really say Google is not a monopolist if they did not offer search engine choice?

Thanks for the response.

The first point makes sense, but it's hard for me to square with breaking up companies. I don't think any of FAANG has played a real censorious role on the internet. I'm just as free to spin up a random webpage to espouse whatever as I was in 1999. People move around the web differently, but I can still send it to whomever I want. I could also see an argument that AMP and similar technologies hurt journalism, but I hardly ever see that claimed.

If it's a question of agency, to me it feels much better to "vote with my feet" and not patronize companies I dislike than to write my congressperson and receive a form response from some intern. You may be right that people just expect better from the internet than federal contracting.

How is your link related to antitrust?

As some variant of millenial, I'm continually bemused by how many of my peers are on board with breaking up "big tech" and use "big tech" products multiple times every single day.

If I bring up federal contracting, which is an industry of comparable size, and which is dominated by companies founded well over 50 years ago [1], and which is way more politically opaque, and whose funding they cannot as tax-paying US citizens opt out of, the response is never anything more than "meh".

I understand that one is more relevant to daily in-their-face life. These aren't dumb people. They read, argue, write, vote. But the choice of what to agitate for is so odd to me.

[1] https://en.wikipedia.org/wiki/Top_100_Contractors_of_the_U.S...

IIRC, one reason his PhD was so hard is he spent a bunch of time working in a grind-it-out field (automated bug-finding) with an advisor who was sort of an outsider in the field. That's a recipe for a ton of hours spent on the wrong stuff with little agency. He only pivoted toward his current focus on HCI, learning, visualization, etc incidentally near the end of his PhD.

Since he's now familiar with and known in his chosen area, he may honestly be able to say to his students "I can help you do meaningful and rewarding research".

(Of course, I dunno what the actual story is.)

They are? Recent history suggests it's way easier for a random person to build a successful search engine (e.g. DuckDuckGo) than a successful telecom, and I have way more choice of search engine than telecom.

You're saying that both "adding search engine choice" and "removing search engine choice" are monopolistic tools for Google?

Not quite a sequel, but he recently gave more details on the post-phd transition here: http://pgbovine.net/one-weird-year.htm.

For anybody interested in the computer science PhD process, a good complement to the featured article is Philip Guo's "The PhD Grind" [1]. It's a long (115p, but with big text) blow-by-blow account of the author's PhD at Stanford in the late 00s. It's neither sensationalized nor sanitized, and it's weirdly engrossing for an account of PhD research. I read it before starting grad school and I think it helped my "meta" PhD skills a lot.

[1] http://pgbovine.net/PhD-memoir/pguo-PhD-grind.pdf

There's a bit of a counterpoint to this: with papers, it's easier to point to your body of work (peer-reviewed and, if you're good/lucky, peer-understood-and-approved) as clear evidence that you can do research. Without papers, or with very few papers, that evidence has to come from somewhere else. My understanding is that somewhere else is usually recommendations, from your advisor and other people in the field. Some reasons to dislike that system include:

1) it gives advisors a lot of influence over the student's career (obviously, if you have an actively hostile advisor things will be hard in any system, but I have known successful faculty who had indifferent fund-and-forget-style advisors who wouldn't know enough about the student's field to write a good recommendation)

2) informal "I like this person, hire them" networks tend to select for people who are like current participants in the network

I agree that a coarse "do they have at least x papers?" way of evaluating researchers is silly. But it's not crazy to have 10 papers when you graduate. That's about one paper every 6 months in a 5-year program. Six months is a long time, almost a thousand hours of work. If three people are focused on a project, it's pretty plausible to get a cool result in six months.

This style of writing where almost every sentence is its own line is weird to me. It's like writing in all caps -- if you emphasize every single point, then you emphasize none of them. Especially when it's a long post that takes up a bunch of space.

Not only is this differentially private, it's locally differentially private, which is an even stronger privacy definition. It's "local" because the user adds randomness themselves. Generic differential privacy is a weaker definition because it lets whoever's running the algorithm collect raw data and then add randomness somewhere in the computation pipeline to produce privatized outputs.

This kind of example also predates the definition of differential privacy by about 40 years [1], although the motivation is pretty much the same.

[1] https://www.jstor.org/stable/2283137?seq=1

Is there a further breakdown between residents and workers? I didn't see one in the article, but I would guess that resident deaths outnumber worker deaths. That seems like an important distinction.

> “As you can imagine,” Nixon confided in Rogers, “there’s strong feeling that we just shouldn’t, as [Reagan] said, he saw these, as he said, he saw these—” Nixon stammered, choosing his words carefully—“these, uh, these cannibals on television last night, and he says, ‘Christ, they weren’t even wearing shoes, and here the United States is going to submit its fate to that,’ and so forth and so on.”

Imagine hemming and hawing to pick the most judicious word and then deciding it's "cannibals".

I'm also amazed that the author of this article, who clearly holds no special love for Nixon, "was the first director of the Richard Nixon Presidential Library and Museum". Though per Wikipedia, it looks like they brought him in to do a specific job:

> He was appointed when control of the Library was transferred from the Richard Nixon Foundation to the National Archives and Records Administration. His biggest task at the library was to present a more objective and unbiased picture of the Watergate scandal—a task completed in March 2011, when the Library's new Watergate gallery opened and received extensive news coverage. Naftali left the Nixon Library later that year.

[1] https://en.wikipedia.org/wiki/Timothy_Naftali

Bonkers indeed. An interesting point of comparison is this "interview" with Prince, written over the course of months by the writer tapped to write his memoirs [1]. It seems that Prince and Little Richard both had an extravagant overflowing creative instinct that just...dazzled people. As someone whose personality tacks way more toward clear and consistent structured logic, it's interesting to see.

[1] https://www.newyorker.com/magazine/2019/09/09/the-book-of-pr...

The study in question -- outcomes for new vs. "refurbished" pacemakers and defibrillators -- sounds like a useful question. Plus, it sounds like he did a bunch of manual work matching each refurbished device use case to similar new device use cases. So kudos to him for sure.

But it's worth noting that "[h]e was aided by his father, MHI [Montreal Heart Institute] cardiologist Dr. Paul Khairy", which helps explain why he could start rotating through groups there when he was 12. That's not to diminish his accomplishment. It's not like the NEJM accepted the paper because it's a cute story. It's more like he had a great opportunity and took advantage of it.

I read this and some of your other posts. As a native English speaker who's read many technical papers, I think your English is totally fine, full stop. Moreover, it's far from being any kind of barrier for technical writing, where the fine points of sentence construction, grammar, flow, etc. are much less important than a clear exposition of ideas.

> If search engine does its job well

This is the problem of blogspam. There are plenty of "popular" blogs that are good at SEO and produce little to no good content.

What are these areas where "you truly have to be a genius to do it"? (Asking for a friend, I assure you.)

+1. His conversation with Neal Stephenson is a fun entry point. A recording and transcript is here: https://medium.com/conversations-with-tyler/tyler-cowen-neal...