kulkarnic

Treasuries, MBS, most non-corporate bonds all trade in eighths or 64ths. Coupon rates are in eighths as well.

Commodities I think no longer do, but did until recently.

Isn't the risk profile of a checking account "don't lose money"?

Yes, that is what I meant. Thank you for expressing it more clearly!

Passive investing isn't zero sum - it's positive sum. If you could buy a fraction of earnings from every business in the economy (i.e. both businesses that currently exist, and future businesses that are founded in the future), then you get a rate of return that is roughly the growth in GDP.

Concentrated portfolios are also positive-sum, and have returns higher than passive investing if you are smart or lucky.

> When we first started investing, we approached it from two beliefs: 1) you are unlikely to grow a portfolio without a small percentage of it allocated to more active investments

I wish you all success, but this assumption goes against about a half-century of academic research. You might say "but it's crypto!" But the law of averages is brutal, and it is agnostic to whether we're in a crypto world or not -- if some fraction of market participants get an above-average return, mathematically, some must get a return that is below-average.

Possibly -- you sell stock in fossil fuel companies, and reinvest in some other, hopefully more sustainable, company. It's not the divestment part that helps, it is the reinvestment.

Yes, the story is inaccurate: the study was about how power, not luck, changes perceptions. But the message is still right. It's easy to be a jerk if you are powerful. (And it's easy to be negative when you're just commenting on the Internet.)

If you actually care about the facts, read the original researcher's essay: https://hbr.org/2016/10/dont-let-power-corrupt-you And if you want to still believe your success is the exclusive result of your hard work, read about Raj Chetty's work on just how much luck matters. http://scopeblog.stanford.edu/2016/04/12/stanford-study-shed...

Edit: I got it wrong, the original study was four people as Lewis reports, not five as some commenters here believe. Lewis is also right about table manners, btw.

This is factually incorrect, in addition to being bad advice. You take on debt when the net present utility of the purchase is greater than the price of the debt.

Here's a simple example to illustrate: you have an interview tomorrow, and you want to buy a suit to be sure you are appropriately dressed for it. The net-present-utility of the suit is quite high -- not having one may cost you the job opportunity. On the other hand, a suit is a depreciating asset. It is also not an income-generating asset (i.e. you may sell the suit the day after the interview at no loss of income.)

Using the reasoning of "debt for appreciating/income generating purchase" will preclude you from buying this suit, sensible though the purchase is.

PS: in an ideal world, you'd rent the suit, or any asset that has only temporary utility and pay depreciation+premium. http://john-joseph-horton.com/papers/sharing.pdf

Before you buy into the negativity of some comments on this thread, take a moment to pause. Andrew has achieved some truly remarkable feats. Why not accept what he has achieved is many standard deviations away from the average, and try and learn from what he thinks was useful?

To me, that the top comment right now is about how Baidu "cheated" on an AI benchmark says both that no one can have perfect oversight, but also that no matter your other achievements, someone will always point out a shortcoming.

I just checked their site, and they act less like a hungry startup, and more like a don't-care near-monopoly.

Three examples:

- You can't see prices for many items without logging in. For a site that is geared towards price-conscious consumers, this is such a silly move. Alternatively, they consider login/registration as a worthwhile conversion (otherwise they could show you the price when you add something to cart.)

- Yet, their registration page doesn't let you create an account with Facebook, Google or anything else. If you care about conversions, act like you care!

- Their Help is an email address or a phone number. Seriously? No chat? No FAQ? Even Comcast has this figured out. If your LTCV is > $500, there is no reason to skimp on support when you don't have traction.

I misphrased above. The index fund will pay out a dividend if the underlying stocks do. Fortunately for investors, Companies are paying dividends less and less frequently, preferring to do stock buybacks (which improve the stock value) instead. Dividends are also "qualified" if you've held the stock long enough, so you pay the lower capgains rate.

Articles like this are evidence that people are still surprised that capital grows faster than labour. At least in fair capitalist society the surest way to become wealthy (statistically speaking) is to understand:

- The total wealth of the world is expected to keep increasing - Capital grows faster than labour - Taxes and disasters redistribute wealth

This is why investing in index funds that don't pay dividends is such a good idea.

You only pay capital gains taxes when you realize gains, not every year.

Trump would pay 15% of $20B = ~$3B, so his net worth would be $17B. And this is assuming that his stock investments don't pay qualified dividends that are reinvested. So, yeah, he'd have much more than $12B.

Yes, you can save yourself from the fallout (which decays exponentially with distance anyway). Unfortunately within around 15mi, you're still going to get pretty bad burns, unless you duck into your shelter immediately after the explosion. That's where the early warning systems are useful-- you get 6-8min with an ICBM strike.

The rational reason we don't build nuclear shelters anymore is because they're no longer effective. As weapon yields increased, it's become apparent that a concrete, underground hideout is not going to save you.

The irrational reason is that it's been 60+ years since atomic weapons were deployed, and we are confident the danger has passed. But really, there's plenty of weapons out there: http://www.motherjones.com/politics/2011/11/map-nuclear-bomb...

Developers need hardware too!

Is this a serious post, or are you being sarcastic? I honestly can't tell. JS the language of the future? Why? It has probably the worst gotchas of any language I've coded in, it's verbose, and weird scoping. Nor is it especially nice to optimize for/with.

I can certainly see being stuck with Javascript (just like we're stuck with the x86 instruction set even if simpler alternatives exist), but I'm not sure it's something I rejoice about. Javascript is like anti-Batman: a language we all deserve, but not one we need.

The prize money is definitely a great idea. It'll probably be seen as a rich-company-PR move, but in hindsight, I think we'll see it as a sign of the times.

Computer science seems to be going through a similar transformation as Palo Alto. Just as Palo Alto went from a place where people were modest about their wealth to one where vanity license-plates decorate Teslas at the curb-side, computer scientists seem to have transformed in public imagination from quiet nerds to celebrity saviors of mankind. I'm not sure the image is well-deserved, but so it goes.

I think the clustering either has been trained very well, or uses a very conservative rule-based system. It seems to work great!

I can't edit this original post, but it seems I can't invite people yet; will update when I do.

I wrote a review of the Inbox on facebook (that was my deal with the person that invited me). Cross-posting here if it interests you.

My review of Google's new Inbox, based on two hours of use (thanks [redacted] for the invite).

First the good parts: 1. I really like bundles. The idea that I can "Sweep" all the promotional emails I get in one click fills me with glee (and marketers with anxiety, I imagine). It's also nice to have all the travel related emails in one place. Never again am I going to be confused whether I should be going to SFO or SJC.

2. I like the new compose. My emails are too long (both those I write and those you do), and I'm praying that showing just one line to write a response will make emails briefer.

3. Snooze/procrastinate: How I like the idea of "someday". Combined with my own future discounting, I'm not going to feel guilty about not responding anymore.

The not-so-good parts:

1. Boy is this opinionated software! There is an inexorable push to empty your inbox. I guess I'll like it if I get with the program.

2. It's too pretty. No, seriously. The title-bar is too bright, there are too many people's faces, too many colors and font styles. I like my email drab so I can focus on what people are saying and not get distracted by the colors.

That's all I see with two hours of use. Oh, also, I don't know how to invite people yet. If someone tells me, I'm happy to invite y'all.

edit: formatting.

Actually, I just returned my iPhone 6 last week, and got a cheapo Moto G with LTE. Yes, it was too big for my hands, and I like using phones with one hand.

How about asking people for the command line arguments once they double-click?

Telling users they are wrong seldom works in the long run, because this requires a change in the mental model. The program model most people using GUIs assume is "you open a program, then you tell it what to do". The CLI model of "you first decide what the program must do, then you run it" just requires more thinking. As Steve Krug points out, "don't make me think" :)

As someone who's from India, but lives in the Palo Alto area, this is an especially astute comment. America's still clearly the innovation leader in many markets, and I think it's fair to expect that many young people (if not governments) look to these corporations to do what's right.

It's hard to know what's right, especially in a global marketplace. But this is certainly a good option.

```The methane is emanating from at least 570 locations, called seeps, from near Cape Hatteras, N.C., to the Georges Bank southeast of Nantucket, Mass. While the seepage is widespread, the researchers estimated that the amount of gas was tiny compared with the amount released from all sources each year. ```

Somewhat sensationalist given the source. I really hope articles like these do not distract from the larger climate change problem.

There might eventually be an API available. Right now, we're focused on getting the actual grading interactions for the peers right. Richard's etcml already has an API.

Um, hey, I'm the author of the paper. I'll check this thread again in a few hours, in case you have questions about it.

This is both amazing and scary. The amazing part is obvious.

The scary part is we now have companies that are worth a billion dollars that are unlisted and have no public scrutiny. Either a billion is not a very large number to the American people today, or we believe that public scrutiny is unnecessary for large companies today. I don't think at all that Stripe is one of them, but wouldn't a billion dollar company be a powerful entity that could invest to subvert the public interest?

That is unnecessarily harsh. Docker legitimately represents a new idea, and some claim, a better idea. Yes, it won't run on RHEL5; but Java probably did not run on DEC mainframes either (back when its selling point was Write once run anywhere).

The question seems to be this: does this container model represent a step forward? I think it does. (And is it hyped a bit too much? Probably :-)

A few days? People get upset if you haven't replied in an hour.

Seriously though, long periods of solitude is not what we need. I'd not be productive if I became a hermit, but I'd be a lot more productive if I had periods of quiet time, punctuated by moments of discussion. This is how thinkers and artists (and hackers) worked for a long time, before we wrongly convinced ourselves that being interruptible improves productivity.