sendos

Possibly. Even if this incarnation is a fad and goes away, it clearly touched a nerve and peaked many people's curiosity in ways other AI agents haven't so far. (I'm talking about broad-use agents, not coding agents)

I don't think gymnastics, shower ads, or medical examinations show naked breasts bouncing, among other things. The models could learn how things might behave from observing clothed bodies, but I'm not sure that's realistic to expect from the models.

Honest question: if you and your kids use YouTube to the degree you describe, why not pay for the ad free experience? Why go through so many hoops? Is it the amount of money, is it the principle, or something else?

Does this need any regulatory approval, or will it go through as is?

People are mentioning this is due to lack of testing, but if that were the case wouldn’t hospitals be overwhelmed like they have been in Italy and elsewhere?

Testing or no testing if hospitals run out of ventilators for people in need, or out of masks, that would still happen if the disease was spreading like it does elsewhere

I agree that if the attitudes of people in Japan bother him so much, not clear why he’s staying

Having said that, the phenomenon he describes is pure racism. Not easy to shrug off

Not a dupe. It’s focusing on the leak instead of the sunsetting

Relevant: http://andrewoneverything.com/post/43840346900/we-will-alway...

I don’t think there will ever be a way to make a “big decision” with thorough analysis that will make you certain you made the right choice. There are too many variables to ever know for sure the “right” choice

A very poignant if cynical poem by the Greek poet Kavafis on this topic:

You said: “I’ll go to another country, go to another shore, find another city better than this one.

Whatever I try to do is fated to turn out wrong and my heart lies buried as though it were something dead. How long can I let my mind moulder in this place? Wherever I turn, wherever I happen to look, I see the black ruins of my life, here, where I’ve spent so many years, wasted them, destroyed them totally.”

You won’t find a new country, won’t find another shore. This city will always pursue you. You will walk the same streets, grow old in the same neighborhoods, will turn gray in these same houses.

You will always end up in this city. Don’t hope for things elsewhere: there is no ship for you, there is no road.

As you’ve wasted your life here, in this small corner, you’ve destroyed it everywhere else in the world.

This is a library that provides access to Matlab-like plotting functions for easily creating javascript plots from within Python.

No need to deal with HTML, Javascript, or CSS to get some plots with multiple linestyles, colors, markers, and labels. Just plain Python code and a simple plotting syntax. The currently supported Javascript plotting libraries are Flot and NVD3 (which is a wrapper around D3).

One use case for this library is to have Python scripts running automatically and updating websites that serve Javascript plots to visitors.

For example, we used this library to create and update the plots on prespredict.com. During US presidential election seasons, a cron job calls the plotting script daily, which automatically updates the plots on the website with data from the latest polls.

The resulting scripts may not be "Pythonic" enough, but I was aiming for the simplicity of creating plots that Matlab has.

I did look at some empirical data to derive the shape of the curve, though the derivation is not as thorough as it can be. If I get time to make the derivation more formal, I'll post in in the methodology section.

One difference I have with 538 is that, I believe, they have some model of the state results, maybe with some correlation, and then simulate the election many times to come up with the PDF of results.

On the other hand, I assume the state results are independent (conditioned on the latest polling) and so I am able to derive a closed form calculation of the PDF, with no simulations. So the independence assumption makes my calculation easier, though may not be as accurate as a model with correlation between states. However, the model has done quite well for the past three elections, so maybe the independence assumption is not that bad.

FYI, the ranking has changed. The results depend on how the lines of the form "1, 10, BLANK, BLANK, 2.04E+07" are interpreted.

Physicians are still up there, but some other occupations have also entered, such as "Farmers, Ranchers"

I sorted the data by which occupation marries within the same occupation the most:

     13.9%    Physicians and Surgeons
     10.7%    Textile Winding, Twisting, and Drawing Out Machine Setters, Operators, and Tenders
     10.1%    none
      9.7%    Farmers, Ranchers, and Other Agricultural Managers
      9.7%    Lawyers, and judges, magistrates, and other judicial workers
      9.5%    Miscellaneous Personal Appearance Workers
      9.3%    Veterinarians
      8.8%    Dentists
      8.5%    Miscellaneous agricultural workers including animal breeders
      8.4%    Postsecondary Teachers
      7.6%    Software Developers, Applications and Systems Software
      7.6%    Health Diagnosing and Treating Practitioners, All Other
      7.5%    Optometrists
      7.3%    Chiropractors
      7.2%    Pharmacists
      6.7%    Elementary and Middle School Teachers
      6.4%    Food Service Managers
      6.2%    Agricultural and Food Scientists
      6.1%    Physical Therapists
      6.1%    Gaming Services Workers
      6.0%    Upholsterers
      5.9%    Communications Equipment Operators, All Other
      5.8%    Air Traffic Controllers and Airfield Operations Specialists
      5.8%    Physical Scientists, All Other
      5.8%    Nurse Anesthetists
      5.5%    Chief executives and legislators
      5.5%    Real Estate Brokers and Sales Agents
      5.4%    Clergy
      5.2%    Marine Engineers and Naval Architects
      5.0%    Psychologists
      4.9%    Lodging Managers
      4.8%    First-Line Supervisors of Retail Sales Workers
      4.7%    Miscellaneous Managers, Including Funeral Service Managers and Postmasters and Mail Superintendents
      4.7%    Medical Scientists, and Life Scientists, All Other
      4.6%    Secondary School Teachers
      4.0%    Textile Knitting and Weaving Machine Setters, Operators, and Tenders
      4.0%    Podiatrists
      4.0%    News Analysts, Reporters and Correspondents
      4.0%    Sewing Machine Operators
      3.9%    Bailiffs, Correctional Officers, and Jailers
      3.9%    First-Line Supervisors of Personal Service Workers
      3.8%    Tailors, Dressmakers, and Sewers
      3.8%    Economists
      3.8%    Musicians, Singers, and Related Workers
      3.8%    Environmental Scientists and Geoscientists
      3.8%    Property, Real Estate, and Community Association Managers
      3.7%    Insurance Sales Agents
      3.5%    Agricultural Inspectors
      3.3%    Butchers and Other Meat, Poultry, and Fish Processing Workers
      3.2%    Morticians, Undertakers, and Funeral Directors

I noticed them also, but they are not the one I had in mind. (what3words does have an interesting approach, I wonder if they are getting any traction)

Results are interesting, though unfortunately not too surprising.

As an aside, I did really well in drawing the curve. The website asked me "Is that you, Raj Chetty?" [http://imgur.com/XU4UNya]

Great product (at least from an ingenuity point of view; would be interesting to see how kids react to it)

Side question: What is this style of info presentation called? Basically, a full-page-width webpage, with very sparse information that slowly shows up as you scroll down. I'm not too fond of it, but I see it a lot lately, so I assume it has been shown to lead to higher conversion rates. Does this presentation style have a name?

A recent thread (https://news.ycombinator.com/item?id=7430501, "What if you didn't need money or attention?") reminded me of this post I wrote a while back, and thought you guys might enjoy.

I'm sure he's a remarkable person, but not so sure about the examples you give as concrete achievements. First, I would not consider Loopt a success. Second, can you give examples of startups he helped succeed?

There is a saying to the effect of "If your financial advisor is so great, where is his yacht?"

Similarly, how can a person be a great startup advisor without significant successes under his belt? If he is that good at strategy and finding product/market fit, why doesn't he apply these skills to his own ideas?

The basic lesson out of this may be that, even if you are a remarkable person, and have great insights and great strategic skills, finding success in the startup world has such a big element of luck and "being in the right place at the right time" that a lot of remarkable people never find success with any startups they start.

Given all this over-the-top fawning from Graham, Thiel, etc, one thing comes to mind: "Where's the beef?"

AFAIK, he only has one failed startup to his name, and yet he is hailed as a shrewd thinker and top startup strategist.

For those who know and respect Sam, how would you reconcile his amazing status/respect-from-top-people with his lack of concrete achievements?

I have a term for this: "etime" [1] (the product of time and mental energy). Both are necessary to get anything done.

[1] http://andrewoneverything.com/post/43544648479/i-dont-have-e...

Does OSX Safari share the same code base as iOS Safari? If yes, why is OSX Safari better able to render websites and better able to work around JS bugs?

If it's due to more SW bugs in iPads than in laptops, can this be due to iOS being relatively immature as an OS compared to OSX and Windows?

Or are there some HW restrictions in tablets that make the SW more difficult to be made stable?

> I'd bet good money that you have a compromised or vulnerable mailer script on your machine somewhere

This must be the case. I turned on SPF and DKIM and the spam emails slowed down for a while, but are now back in full force.

It looks like I will need to torch the whole VPS (I have backed up the important data)

I call the effect smartphones are having on us "Information Obesity" [1].

Just like when food was scarce and people ate as much as they could, and now that food is abundant many people can't limit themselves to sensible amounts, a similar thing happened with information.

When information was scarce, we consumed as much as we could, and now that it is abundant many people can't limit themselves to sensible amounts, and keep checking their email, twitter, reddit, HN, etc, on their smartphones.

Just as we learned to teach people who to deal with food abundance (and not successfully yet, unfortunately) we should one day start to teach people who to deal with information abundance

[1] http://andrewoneverything.com/post/43839194291/information-o...

I did not follow this case when it broke, so do guys know what Aaron planned on doing with all that JSTOR material he was downloading? (~80% of JSTOR, according to one article I read)

"Does that reflect an injustice? Many say no, arguing that women earn less because they take time off to care for children or elderly parents (and thus have less seniority) or accept lower pay in return for more flexible working arrangements. Women also tend to go into lower-paying lines of work, shunning higher-paying technical fields. It's easy to caricature this view (dirndls versus Visigoths, etc), but there may be some truth in it. Some research suggests that when women behave as men do--not having babies, mainly--the income gap largely disappears. If so (I won't claim the matter has been definitively settled), the question facing women is a stark one: What do you want, kids or cash?"

[1] http://www.straightdope.com/columns/read/2411/are-women-paid...

> > the Foreign Intelligence Surveillance Court has for years been developing what is effectively a secret and unchallenged body of law on core Fourth Amendment issues, producing lengthy classified rulings based on the arguments of the federal government — the only party allowed in the courtroom

It seems to me that we are doing a disservice to the terms 'law' and 'court' if we call a "secret and unchallenged body" of decisions 'law' and we call a place where the federal government is "the only party allowed in" a 'courtroom'

Let's not give this travesty any more legitimacy than it deserves

> It is sad indeed, but perhaps every political system sooner or later becomes a tyranny and dictatorship. It may be just a matter of time.

Quite possibly.

I wrote about this a while back, after watching a PBS documentary on domestic spying. Basically, it seems to me that the US could one day turn into a totalitarian regime with no way back [1]. Maybe it's overly-pessimistic, but once all the total-surveillance tools are in place, if the government ever turns totalitarian, it will be extremely difficult to overturn it

[1] http://andrewoneverything.com/post/43837476217/the-us-could-...

This is related to the recent "Why the Fuck?" thread (http://news.ycombinator.com/item?id=5088815) and is something I wrote a couple of years ago

The following modifies only the file size:

  ls -l | perl -pe 'while(s/^((\S+\s+){4})(\d+)(\d{3})([^\d].*)?$/$1$3,$4$5/){}'