Sunday assorted links

Comments

1 is a great article. I have used IPython Notebook, now Jupyter, for a few things and liked it. One advise for Linux/Mac users is to create a new user for it. Having different versions of Python and libraries on a system, while easier, can still be a pain. Or go whole hog and use a vm to test and explore.

Pretty amazing that "The Cathedral and the Bazaar" was published just 21 years ago. Helluva global impact in that time.

One book I'm reading to learn about machine learning (Frank Kane's) has jupyter notebooks as bonus material. The interactivity has helped me a great deal.

What example of research results would you recommend to a budding student in the field in Africa with a $20 phone?

I could send such a student a link to a PDF if a webpage of a research result was not available, or worst case, send it in text in email, or even tweets.

We can read papers published centuries ago, but can't run programs that papers were written about from a quarter century ago, except at a very good computer museum.

There is much more home grown Africa tech these days: Africode

I don't see any tech in that blog. African or otherwise. It reads like a Creative Writing exercise. And the main author?

Tosin Adeniji is an expert creative marketer, passionate about media tech and impacting young people. She currently works as Senior Product Manager within the Verizon innovation team in New York City and holds a MBA from Cornell Tech. Prior to her Masters, she was in London where she worked within entertainment for companies such as EMI, Universal and LoveLive. Over the years she has worked and consulted for the likes of Spotify, Google and HP.

Not in Africa. Not in tech either. Unless you count the girls who work in IBM's HR department as being in tech.

Jupyter notebooks hosted on a personal web page + sharing data analysis code on Github is probably the best applied academic resume I can think of right now. You're sharing both process and output in an interactive fashion. Open-source is key. The future of professional success is all about that internet karma.

As a Linux user and frequent Python/Jupyter user for STEM research, I recommend all newbs to just stick with Python 3.x. Who's seriously using 2.x right now?

6 - I didn't know that Mark Zuckerberg and I had a fiduciary "(prudent man" rule and all that) relationship. However, I have a problem with him in 2012 allowing Obama's agents to use my personal data.

FYI - My personal data (some of it fictitious) is worth exactly what Mr. Z. paid for it.

Seriously, the problem is not one of fiduciary trust between me and Zuck. It is Facebook reputational. Facebook management helped Obama and didn't have a clue about being used by Cambridge Analytical.

2. Solve for the goat equilibrium.
From the article:

Asked about the possible cost of the rescue, Tilson says, "We didn't even calculate it. We were just trying to be a good neighbor and get the goats back safely."

Ok,ok I have the equilibrium, Pennsylvania goes bankrupt.

This is just a test to see if I can post a comment at all. I already know the new format does not allow me to respond to a previous comment.

I must have more pull than you, because I can respond to any comment I want.

He might have external JavaScript disabled. Through experimentation, I find you have to enable 2378nh2nfow32gm3mb25krmuyy.wpengine.netdna-cdn.com to do a nested comment.,

6. Why would Zuckerberg deliberately restrict the data that Facebook holds? What interest would they have in that? Why would they show untargetted ads when they have limited real estate? Yes, he has the power to do that, but what does he gain?

One of the problems with the Cambridge Analytica story is that there's a few people who are obsessed with it, a few doomsayers about mind control, some media publications that see it as a way to clip Facebook's wings and gain back some ad revenue, but the problem is that most of the public, including most of Facebook's users really don't care that much. They aren't obsessive about privacy and targetting.

Laws. Reputation.

Wouldn't the color of the links improve if it was a bit darker? I find it hard to read on desktop

4. Am I the only one to notice that Douthat's essays are like a ride on a roller coaster: up, down, and all around. What Douthat doesn't like about Williamson is that Williamson's honesty doesn't sell. This past Sunday, Easter Sunday, the priest in his homily stated that it was a fact that thousands saw the risen Jesus, and that absent the risen Jesus Christianity was a fraud. Oh, my! According to the Gospel of Mark, the first Gospel, nobody saw the risen Jesus. Did Mark leave it out because he didn't deem it important? Douthat is a fundamentalist. Sure, he claims to be a Roman Catholic. Marco Rubio claims to be a Roman Catholic too, yet he was baptized a Mormon and attends an evangelical Protestant church. Up, down, and all around.

Actually, no Mark records many of the same sights as the other Gospels - Mary Magdalene, two unnamed disciples, and then the eleven disciples as a whole.

As far as it "selling" there are many things it is illegal for me to do as a physician. Part of getting the license means that we take on obligations that the general population does not have (i.e. you cannot be tried for giving false medical information, I can be unless I take great pains to say that I am not speaking as a licensed expert). It is perfectly in keeping with the fact that we also tend not prosecute people who are drug seeking, but will go after pill mill docs who prescribe against the law (e.g. prescribing opiates for "menstrual cramps" to men). Things become negligent for us that merely asking for is not a crime for the general population.

Nor is medicine the only place where we see this sort of one-sided prosecution. Sweden, for instance, has decriminalized the sale of sex while making the purchasing thereof illegal. This appears to better serve the public good (as is traditionally defined with diminishing STI spread, ancillary crime, sex trafficking, and violence against sex workers) than either complete prohibition (victims afraid to come forward) or or complete legalization (looking for needles in haystacks).

Special privileges, special responsibilities. I think historically under common law, you saw this in some of the other professions as well, and perhaps in a few other unusual jobs (e.g. ship captain).

Perhaps someone with a better grounding in common law will correct me if this is not the case.

Tim Kaine claimed to be a Catholic but (like his running mate) didn’t want to enforce any restrictions at all on abortions ( putting him at odds with the majority of the population). And gosh, did you see those 1000+ NYT commenters, completely predictable — if you’re not a woman then you can’t comment on abortion.

Democratic Catholics tend to ignore abortion, just like Republican Catholics tend to ignore the death penalty.

I'm 100% positive that you can come up with a line-drawing exercise to show why one is okay and the other isn't, either way, because I've seen people do it a million times. It's unfortunate that most people argue for "what can I get away with?" instead of "what ought I do?" but that's the world.

I like Douthat a lot, but I wonder if in all his 'reasonableness' he would be able to agree with my take on abortion, where it should only be legal in the first trimester except in extreme cases of fetal abnormality or mother's health risk. He rightly decries the extreme pro-choice position, but he also keeps referring to fetii as people.

As I've said, if you think even a day-old fertilized egg is a person, then that's that, no compromise is possible. I suspect that's Douthat.

So while he's right to criticize the 'abortion anytime' crowd, if he is 'abortion never' then he's an extremist too.

Surely even you must recognize the huge number of logical flaws in that post.

To start with, are the two positions - human from conception and abortion up to and after contractions have started - equivalent? They are only both equally extreme is there is a perfectly symmetrical distribution of views between those two end points. I doubt that is true. I would think that Douhat is closer to the majority of Americans on this. But even if he is not, there is no reason to think that most people are not closer to one than the other.

Nor is it true that if the day-old egg is human no compromise is possible. After all, the stakes are very different. The Catholic end of this argument wants to end what they see as murder. The feminist end of this argument wants life to be more convenient. Those two are not equivalent. It is perfectly possible to compromise on a total abortion ban because abortion just is not that important to the feminist end of the spectrum. It is important to the Catholic end.

It would not be hard to imagine a recasting of abortion as oppressive to women. In which case the people who support it now would oppose it. As feminists used to. If we encouraged women to sue their abortion providers for being callous and exploitative, the pro-abortion Left would be divided but they would probably come down on the side of the women.

Also historically, the Left has had no problems with total abortion bans when Stalin or Mao imposed them. It is not a central issue.

And no one can agree on your view because it is nonsense. This is a debate between people who want abortion up to and including the moment the baby's head is out of the cervix and those that do not want it at all. The West has compromised on laws like yours that enables people like you to feel better about the fact that everyone has abortion up to and including the moment the baby's head comes out on its own. A first trimester ban is a fig leaf. Extremists will continue to do whatever they like and the Left will cover for them. As the mainstream media did all it could to keep Gosnell from the front page. The choices are really up to (and even after for Gosnell) birth or not at all.

After all, would you agree to jail a woman who had an abortion at three months and one day?

And this is what I'm talking about. You're such a blinkered partisan extremist you can only post like this. Douthat at least tries to deal with the middle vs the extremes. Fortunately most of this kind of discourse is just anonymous internet combox stuff.

"SMfS - And no one can agree on your view because it is nonsense."

I don't think what he said was nonsense. Both "Always abortion" and "Never abortion" positions would be far outside the American extreme.

I don't think you'd get even 20% support for allowing abortions at 40 weeks and I don't think you'd get 20% support for No Day 1 abortion even assuming rape.

Of course what I said isn't nonsense, it's the common sense moderate middle. SMFS (and his opposites/analogues on the extreme left) literally cannot see the world that way. Of course it may just be a troll pose. He likes to pronounce extreme silliness as if it were the most obvious truth.

3. "They" still haven't told us what triggered the Salem witch trials.

My guess is the answer for both is the same.

The hysteria that led to the Salem trials: was it mainly a male thing? What about the Havana business?

3) Does the average embassy diplomat usually score at the population average on standardized tests? I would have thought the selection mechanism would lead to them being above average, on average, in which case a 40% cutoff would not be unreasonable at all.

Also, the argument "maybe they imagined it" (the meaning of the final sentence stated in plain terms) is scientifically equivalent to "I dunno ... maybe God? Magic?"

Use of neuro-effective microwave weapons (including with sonic effects via thermoacoustic effects and bone conduction) and monitoring goes back at least as far as the 60s or 70s, with Moscow embassy staff reports of such things being published as early as 1976.

4. The "paper" by "united for life" which i doubt cares about legal persons, is "fake law", ignoring the record of women prosecuted for "murdering their fetus" by taking a drug. These laws have been passed in recent years post Roe and are open ended enough that the most common aboortion method would result in a murder prosecution, especially of poor women and non-white women.

Bei Bei Shuai, for example, prosecuted for murder by Indiana, when the governor was now VP Pence. The law was passed in 1979 in response to a miscarriage after a woman was beaten by an abusive partner.

Alabama has prosecuted about 40 women for "murder of their fetus".

It's clear the abortion drugs would be made available throughout the US by activists, in ways that make it impossible for individual States to prosecute the pill supplier. Federal laws would work as well as the Federal laws to block FedEx, USPS delivering fentanyl from China.

Thus, the only person to charge will be the woman who took the pills to get an abortion.

Could you provide a link for your claim about Alabama prosecutions? A quick google yielded “Ala. Code § 13A-6-1 (2006) defines "person," for the purpose of criminal homicide or assaults, to include an unborn child in utero at any stage of development, regardless of viability and specifies that nothing in the act shall make it a crime to perform or obtain an abortion that is otherwise legal“ so there would seem to be a contradiction here.

Perhaps you are referring to the articles by such neutral papers as Mother Jones, Think Progress, etc about the unintended consequences of Alabama’s meth lab law. As usual the irony is astounding. The most virulent critics of smoking or drinking while pregnant think that taking drugs and running a meth lab while pregnant are OK.

https://www.nbcnews.com/news/asian-america/indiana-has-now-charged-two-asian-american-women-feticide-n332761

Here you go on the Indiana cases. Feel free to pick at the difference between infanticide and a miscarriage or suicide and a murder. They were prosecuted in any event and no reason to think they wouldn't be in other states if the right prosecutor or gov comes along.

1. Papers should be more relatable and more accessible. The news media is a poor interpreter.

4. I counter with https://www.theatlantic.com/politics/archive/2018/04/a-dissent-concerning-kevin-williamson/484052/.

Intellectuals from both parties have stagnated and are now realizing the same old same old will no longer cut it. Yuval Levin once commented that we are headed for a generational clash that will fundamentally alter the parties. Buckle your seats and get ready for the ride.

4. For comedians, the Chapo boys are insightful this week.

https://m.soundcloud.com/chapo-trap-house/episode-200-the-jeffrey-goldberg-variations-3918

2 . Seems like the GOAT rescue. (Greatest Of All Time).

"In this day and age, when things can go terribly wrong," he says, "it was great to see things go right."

Punishment for a woman for getting an abortion could be she has to go on something like Norplant (https://en.wikipedia.org/wiki/Levonorgestrel-releasing_implant). So that in the future she has to make a conscious decision to get pregnant.

Maybe not unreasonable to suggest as a requirement after a second, or perhaps third, abortion.

For some variety of reasons, I disagree about it being reasonable for a single case.

However, the religious basis of anti-abortion views among many people who opposed access to abortion might lead them to oppose it. For example, I doubt that abortion is common among married couples.

#1 -- Oh, please. Mathematica notebooks are unreadable even if you wrote them yourself. (I mean in the sense of a logical train of thought, not figuring out what a given expression means)

When you write a paper, you're trying to be expressive. You're trying to tell your audience your result, not preparing an audit trail for how you got the result. The last thing you want is for the audience to screw around with the equations so that the text no longer matches the output. If they want to do that, they can type the equations into their own copy of Mathematica and go wild.

Mathematica is very useful for rote algebra plug and chug. But you will quickly hit its limits if you want to do a real numerical calculation (the kind of stuff that uses multiple cpus for multiple hours) or a nontrivial derivation.

I've met Fernando Perez, and I like him. I'm a big fan of ipython, as well. But ipython, Jupyter, and Mathematica are functional tools, not expressive ones. For communicating results to your audience, LaTeX and prepared figures are much more appropriate.

Comments for this post are closed