nhstanley

Similar effort by Microsoft: https://github.com/Microsoft/Pyjion

Yup. It's hardly any different to the standard sorting, because facebook's business is controlling what you see. I don't know why people keep spreading that like it's some protip.

I think that sets a bad environment, because who decides what historically important is? How many years does it become "historic"?

A much simpler and better solution is to not extend copyright to 120 years or whatever the hell it is now [1] (Thanks, Disney!). Twenty years or so should be plenty for all artistic works.

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

No, but I think as long as the bug bounties pay enough to keep someone comfortable (along with added notoriety/resume padding with it), you'll have enough moral people choosing to reveal them to the companies rather than bad actors. Maybe that's naive. $1k bucks though probably isn't that number. More like $50k or $100k for 0-day level stuff.

I know this is anecdotal, but everyone I know uses Venmo. I have never met anyone who uses Square Cash or Messenger, and I have both apps. To be fair, I don't know anyone in the US that uses WhatsApp, but all my friends in Europe use it. So, maybe it's a regional thing? I would love to see real usage statistics on those.

Why would they need samples when the chemical structure is so widely known[1]? It's not like it's a biologic where sequence and other formulation details are critically important. Can you explain?

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

Edit: I noticed that paulmd, further down, explained why this is important. I didn't realize the rules were this stupid. The FDA should be able to force companies to make their product available for comparison testing. It's insane that it isn't the case.

Stuff like this often makes me consider writing an open source version of whatever the software is. If I really had to base my long-term livelihood off a piece of proprietary software, I wouldn't want to perpetually have such a sword hanging over my head. I work with many pieces of software where the core functionality is actually quite simple (more complex if you want nice features), and I'm actually doing this for one of them.

Perhaps making the student share a portion of the cost would be smarter. Maybe make the student pay 5 or 10k or the 30.

Alternatively, everyone could be adults and settle for a smaller, much more reasonable amount proportional to harm. Especially considering the company wasn't really harmed substantially by this since he did have access, just not on that machine.

Not my opinion, but what I think lukasm was getting at is that a majority of players in the startup game aren't VCs or founders, but suckers taking low wages and/or options and/or insane hours on the promise of a big future payout that never comes.

"Why did you want to climb Mount Everest?"

"Because it's there. Everest is the highest mountain in the world, and no man has reached its summit. Its existence is a challenge. The answer is instinctive, a part, I suppose, of man's desire to conquer the universe." - George Mallory

"Cancer" is a word that describes any disease that is the result of uncontrolled growth of your own cells. However, there are so many types of cancer cells and the treatment for each so different that saying we've "cured cancer" is a prime reason why the public doesn't understand or trust these miracle breakthroughs.

That being said, these PD-1 inhibitors are a huge breakthrough for any type of cancer that uses this mechanism to evade the immune system. One of the hallmarks of cancer cell growth is their ability to evade the immune system. PD-1 (programmed cell death 1) is a receptor on immune cells like T cells that tells these immune cells to "stand down"/go inactive. Some types of cancer secrete molecules that activate this receptor, inhibiting the immune cells. Drugs like Nivolumab block these receptors, keeping the immune cells active, allowing them to attack cancer cells when they encounter them.

Metastisis (cancer cells spreading around the body) is what kills cancer patients, and drug strategies that can keep this from happening are key to keeping people alive. (Re-)activating the immune system against the cancer cells is a great strategy that's had some recent breakthroughs. Another recent breakthrough method, called CAR-T cell therapy (chimeric antigen receptor t-cells) also activates the immune system against cancer. In that case, they take immune cells out of the cancer patients body, use gene therapy to inject genetic material needed by the cells to recognize cancer cells, then grow those cell and re-inject them back to the body. This has been wildly successful in certain types of cancer, like certain leukemias and lymphomas, and may be successful for other types. Here is a nice article that covers these general strategies for those interested: http://www.forbes.com/sites/matthewherper/2015/04/20/immune-...

There have been a lot of breakthroughs recently. If we keep pushing like this we'll someday soon be fortunate enough to be legitimately surprised when a friend or family member dies of cancer, and many other diseases.

Certainly the leading cause of liver failure and need of liver transplants. http://www.medpagetoday.com/Psychiatry/Depression/2233

I wont disagree that perhaps the US system could use reform/overhaul, but it's a bit more complicated than you describe. When you compare how medical schools and medical school admissions are handled in the US versus how law schools are organized, it's no surprise why law school has a glut of graduates who can't find jobs. I know this because I have enough friend who went on either of those tracks (as well as others). The med school kids all have jobs that pay well-enough, and the law school kids all became disilusioned and moved to other fields (except for the one that went to NYU).

The number of Medical school admissions and future residency positions is pretty well controlled, so this kind of problem doesn't really arise (well, it's complicated but more or less).

On the other hand, any school can spin up a law program and start charging admission, no matter how qualified it is to actually do that. But it's only worthwhile going to law school if you can get into the first tier (Harvard, Yale, Columbia, NYU) or second tier schools. Third and fourth tier are basically a joke. You're throwing your money away. Of course, many people ignore that thinking they'll be different or lucky. And we end up with the kind of issues we see here today.

It's still more complicated, though, because a lot of what people from third and fourth tier schools used to do (sit in giant rooms and review case law for big cases), they are no longer needed for thanks to technology. So the 2000's onward was an awful time to want to be shove yourself through law school hail-mary hoping that just because you're a "lawyer" you'll be OK.

Law schools students got squeezed on both sides, and the pain is real. A profession that used to be much like medicine, where you would do "OK" in life at the very worst is now a commodity product thanks to forces from all sides.

Someone was arguing to me recently that this has all come about because of some court cases in the 80s and 90s. They claim that one case in particular, between Notre Dame and the faculty, was a turning point.

The story has been related to me like this: Traditionally universities were heavily influenced by the faculty, who had big sway over big decisions. This all came to a head in the 80s when faculty and the administration disagreed over some big issue (like who should be the next president, I believe). The faculty and administration went to court, and the court ruled that the faculty were employees and the administration, essentially, had absolute power and could appoint who they wanted.

This precedent drastically changed how universities were run, and since then faculty has basically been eviscerated (both from a power standpoint and a pay standpoint, ala adjuncts) and universities are now run like businesses (top down management vs. a more distributed/democratic form previously). I haven't been able to find the court case or citations to back this up (closest is a book[1]), but it wouldn't surprise me if true. It would explain a lot that has happened, and if you combine it with other factors like government backed student loans, it's no surprise universities are cash cows run by feudal lords.

[1] https://books.google.es/books?id=PiPktdWvGvkC

What's so entitled about it? The IMF is a creditor just like any other, and assumes risks when it makes loans. The duty of the Greek government is—without question—to its people. While you can certainly argue that default is not in the people's best interest, if the leaders honestly think that's the best option then that's what they should do.

This isn't the first antibody therapy that tries to reduce plaques (Bapineuzumab, solanezumab, gantenerumab, crenezumab, etc.), but is it the first one to "work" correctly, but still fail? I don't believe so. I think bapineuzumab also showed early promise at reducing plaques, though I don't remember if they did a cognitive analysis that Biogen did here.

Don't get me wrong. I think this is the most promising one we've had, if for no other reason than the method used to develop it. They took people who lived into old age without Alzheimer's, and found antibodies from them that targeted Abeta (so they claim). Really cool stuff, if true, and very promising.

$100k in SF is laughable (for that kind of commitment). If you're not offering $200k minimum, it's not even worth 60 seconds of consideration. Unless you really really believe in that 6%. But man you really shouldn't.

Edit: To back this up a little more:

To buy a median house in SF you need to make ~$142k.

http://www.washingtonpost.com/blogs/wonkblog/wp/2015/03/08/m...

The 95th percentile in income is WAY higher in SF than anywhere else, and so is the cost of living.

https://twitter.com/conradhackett/status/578004292151341056

My opinion is that if you're not clearly putting me in a comfortable position financially, why would I ever make such a sacrifice?

> Yet Mr Buffett’s 50th letter to shareholders is an exception, serving to muddy rather than clarify, for two reasons. First, because it does not tackle the questions that hang over Berkshire’s conglomerate model and its durability. Second, because of the uncharacteristic coyness with which Mr Buffett and his partner Charlie Munger —respectively aged 84 and 91—discuss how or when they will give up their jobs.

Did this guy not read the letter? Because Buffet spends a lot of time discussing why he thinks his version of the conglomerate model works and will continue to work. You can disagree with his reasoning, but he certainly discusses it at length.

Second, there have been rumors that Buffet is planning to give up control soon, perhaps at this shareholder meeting. Even if he doesn't, it's not clear that it matters. The final word on asset allocation may come from him, but he is surrounded by people essentially making those decisions for him (as he basically implies in the letter).

Nowhere in this do I get the impression this guy actually read the FULL letter, beyond the first part (meaning beyond page 21).

Was the use of ANN necessary, or simply done because it was cool to do? Because at least in my mind it looks like extracting the text and running a brute force program on it would be faster, no?

How frequently do those people move to those universities after they've made the scientific breakthrough for which they receive the award?

I don't know what the truth is either, but sites like Glassdoor are specifically designed so you can find out if you're being underpaid, and let others know they're being underpaid. So intuitively it could go the opposite as well. People who think they're getting underpaid report, and people who are clearly not ($200k+ crowd) may not. It would be really interesting to know, though.

What would "Extra Discretionary" look like for SF?

Empathy for whom? I know pacifists want to believe we live in a post-war world, and if only we would not make guns or do the kind of things the NSA is doing here we'll suddenly live in a perfect world—but by what fantastical mechanism do you actually think that will happen?

I don't have as much personal experience (only one colossal flop, $4k invested + ~20% lower salary), but I've got enough experience in the industry to have come up with this hard and fast rule: If you didn't found the company and aren't getting 10%+, equity is meaningless. If they insist you take equity ("Don't you have faith that we'll be successful?"), you can offer to knock off 1% of what you would have expected in annual salary without it (even that is arguably too much). Otherwise it's absolutely not worth it.

Companies love trying to get you to take a 5-40% pay cut because "It's a startup" and "we'll give you equity". Total BS. You will be better off spending an hour every week minding your money in other ways, with your appropriate salary.

When you think about it, Germany is like the hot girl in a small town high school. She doesn't care if she's pretty (rich) in the absolute sense, she only cares that she's prettier (richer) than all the other girls in the small town (Europe).

Instead of focusing on optimizing Europe's collective wealth, they are trying to maximize their wealth relative to the rest. There is certainly concerns about things like corruption in the periphery, and Spain and Greece certainly had their inefficiencies, but Germany is going about fixing it in a completely wrong way.

> statistical facts

Didn't downvote you, but you have heard the saying "Lies, damned lies, and statistics", haven't you? Calling a single chart "statistical facts" is laughable at best. Especially when we've seen two versions of that chart (see zaroth's post), with a near identical title (one has "US families", the other "US households") and claiming the same source data from the same blog, showing two very different plots. (Looking more carefully, the trends at least seem similar, so perhaps it is just an issue with the definition of "family" vs. "household"?)

Further, it's pretty clear that simply changing the brackets could result in a very different plot. And that is what we see in the NY Times article, which has the "households" plot but with the income brackets changed slightly and the 2009-2013 data added. In the NYTimes plot it looks like, since 2000, the top bracket has peaked and shrunk, the middle has continued to shrink, and the lower grew. It's a short trend compared to the full graph, but is that the direction we're heading now? So at best, this is all highly subjective, and a perfect example of when "statistical facts" can be manipulated to serve and agenda, populist or otherwise.

In short, nothing was "completely debunked" here, and a great example of why different sides of a debate end up talking past one another instead of acknowledging how complex these things can be. Two groups with different starting positions looked at the same or very similar plots and drew different conclusions. Color me shocked.

Well, in the article the make clear that there is case law that this is illegal and considered "misappropriation" insider trading.[1]

Even if we decide it isn't insider trading, Capital One is on the hook for what it's employees do. Chipotle or whoever could sue CapOne for breach of contract, and CapOne could sue these guys for that. So even if we decide as a society it's not criminal, another group of guys doing the same thing somewhere else might not get to keep all that sweet dough they made purely on civil grounds. I don't know. It's interesting.

[1] https://supreme.justia.com/cases/federal/us/521/642/case.htm...

While there may not be any private beach under CA law, Kholsa actually tried to use a sneaky legal argument to get around that[1]. Essentially, he argued that federal land patent on the property from the 1800s superseded the California Coastal Act. It didn't work, but had it work it would have set a dangerous precedent for any properties that could make a similar claim.

[1] http://www.mercurynews.com/san-mateo-county-times/ci_2614203...

I think the two desk idea is great, though I can see it being an issue when space is limited.

Naive follow up question: Can doing something like routing through a VPN help with this? I assume not, since a forced cookie on your browser will make you always identifiable to the person on the other end, correct?