chris_va

36 ms·

481.
by chris_va 10y ago
Nice to meet you too :)
482.
by chris_va 10y ago
You perhaps underestimate the complexity and fidelity of the Google crawl, indexing, and ranking pipeline. Many intelligent people have spent many years trying to solve the sorts of problems you describe. (disclaimer: I work at Google, and
483.
by chris_va 10y ago
I think your explanation on the racist name calling shows why this is such a difficult topic. There is a incorrect assumption that people often make, without malice, about economics being a zero sum game. It really is not, almost by definit
484.
by chris_va 10y ago
Just a reminder that the Iceland study result was unexpected, and while it is great that the media is publicizing the results, it would be nice to get additional confirmations before getting really excited. Having said that, I'm still
485.
by chris_va 10y ago
This site has estimates for agricultural workforce over time: https://ourworldindata.org/agricultural-employment/ It is impressive to see how small the fraction has gotten already.
486.
by chris_va 10y ago
> And yet people starve all over the place; not just in 3rd world countries but here in the USA. So either it's not that plentiful or we have a serious logistical problem. It is the latter, in case you were curious. Access to food r
487.
by chris_va 10y ago
Many drivers will work for multiple companies in parallel, e.g. Lyft and Uber at the same time. Even if they only get 10% of their rides from Lyft, it isn't a huge cost to the driver. As a result, from the driver perspective, the netwo
488.
by chris_va 10y ago
"No one in the court system is held responsible." ... That is not entirely true (not to say it is false, but more nuanced). Much of the court system is elected (or directly appointed by someone elected), which is supposed to balan
489.
by chris_va 10y ago
I remember the story ending with "they sent back a picture of handcuffs, and so I sent the actual money"
490.
by chris_va 10y ago
Though 2^n could be much larger than the number of items in the observable universe fairly quickly.
491.
by chris_va 10y ago
There isn't much money in news content, at least by internet giant metrics, so I think its success is somewhat irrelevant.
492.
by chris_va 10y ago
To me, SF does not seem particularly walkable, bike friendly, nor transit friendly (it is ranked #2 for all). It isn't terrible, but #2 in the nation?
493.
by chris_va 10y ago
I think this analogy breaks down with ridesharing services because there isn't any lock-in, and barrier to entry is low (now that Uber has plowed through the regulations). Many Uber drivers are already logged into Lyft at the same time
494.
by chris_va 10y ago
A carbon tax would solve the problem. Actually CCS is only about 15% more expensive, so it wouldn't even be that painful. Pumped hydro works really well, 85% efficiency, but there aren't .any places to put it.
495.
by chris_va 10y ago
Hi, feedback noted. I had not thought my comment would generate so much discussion, so yes I should have linked to a more up-to-date source (actually I thought I was linking to the 2015 data, and this was the easiest to read chart, but oops
496.
by chris_va 10y ago
Renewable energy accounting is shady (no pun intended). This is more accurate (original source EIA): http://www.safremaenergy.com/wp-content/uploads/2013/11/LCOE... ... and also why natural gas will alwa
497.
by chris_va 10y ago
Out of curiosity, what would the tech stack look like? What I was picturing was: - LCD display - Patterned IR image projected onto eye - Camera to read IR image on the back of retina, to figure out current focal length of cornea - Realtime
498.
by chris_va 10y ago
You could approximate this using a standard display if you could dynamically track and respond to focal depth. That might be harder than getting a functional light field display, though.
499.
by chris_va 10y ago
Interestingly, the couple might not be worth anything near $400M, and they haven't yet cracked the trust. If I had to guess, I'd assume that everything was bought on credit, and the true net worth is somewhere south of $40M.
500.
by chris_va 10y ago
Actually, maybe not, if you have a large dataset that you are willing to share.
501.
by chris_va 10y ago
Technically that would be up to congress, since the executive branch does not set budgetary policy. Of course, they approve it, so YMMV.
502.
by chris_va 10y ago
Not to belittle the annoyance of the visa/green card process, but you do get to stay in the country unlike the group this is targeting. I think this brings startup founders closer to parity with corporate sponsored visas.
503.
by chris_va 10y ago
I'm impressed that they were able to find 24 references to cite.
504.
by chris_va 10y ago
(Disclaimer: Google engineer and former Google News TL from a while back) Honestly, it's really difficult. It would be great if something like Google News worked (and it does generate a diverse set of biases), but it is just aggregatin
505.
by chris_va 10y ago
For a CPU, you still need to address the transistors somehow. Chips aren't pure transistors, there are wires coming down from above, and that takes up a lot of space. As a result, vertical density doesn't scale as you might hope.
506.
by chris_va 10y ago
I think that training on MD trajectories may be a mistake. Basically they are just creating a deep learning encoder that learns the MD weights, which are often notoriously inaccurate. I may be reading this wrong, though.
507.
by chris_va 10y ago
In the "picture is worth 1000 words" category: http://mobile.journals.lww.com/transplantationdirect/_layout... ... Kind of begs the question about full specimen cryogenic hibernation.
508.
by chris_va 10y ago
Slightly better discussion links: http://news.columbia.edu/content/Columbia-Chemists-Find-Key-... http://science.sciencemag.org/content/353/6306/1365.full
509.
by chris_va 10y ago
Selling IT to hospitals is very very very painful, just an FYI for anyone here thinking about this line of work.
510.
by chris_va 10y ago
Did you mean diffentiating tissue? Yes, that is true. You can use ISO levels from the images to segment different density regions, which is more or less what a doctor would do, AFAIK. Usually the entire picture is non-zero, so you have to s

More ›