HNHacker News
TopNewBestAskShowJobs

gburt

1,600 karma · joined May 8, 2011

submissionscomments
gburt··on Tell dang/HN: <> Title tags shouldn't be escaped twice
For the record, this has been fixed.
gburt··on dang, can we get a collapse threads function?
I use this. Works very well. It is simple enough I'd encourage building it in to the site though.
gburt··on Indonesia's palm oil fires emitted more greenhouse gases in a day than the U.S.
What he is saying is that it doesn't grow. It is a constant relationship. Cows sequester GHGs from the atmosphere (via the food they eat, mostly) and then release it.

I don't have enough data to know if I agree with his point (especially since the GHG changes form), but his "long-term" refers to the life cycle of a single cow.

gburt··on The Lost History of Gay Adult Adoption
This reminds me of the tiny house movement that avoids building regulations w.r.t. minimize size and grid hookup by putting wheels on their semi-permanent houses.
gburt··on Simple Sequential A/B Testing
You would need to propose a method to compare the confidence intervals of A1, A2 and B1, B2 without hurting our target coverage probability on the outer confidence interval. This is starting to sound really complicated.
gburt··on Facebook paid £4,327 corporation tax in the UK in 2014
The part I don't understand is that the brand was developed in the US. Don't you have to transfer the IP at something approaching fair market value to the tax haven corporation? How could you possibly have the resources (in the haven, no less) to do this without loaning money (at interest?) backwards?
gburt··on Global eradication of wild poliovirus type 2 declared
Good point. My use of the word "strongest" was definitely misleading, but I am not convinced it is always the case that there isn't a relatively low dimensional definition of "strong" in this sense.
gburt··on Global eradication of wild poliovirus type 2 declared
This is a clear win for "public health", leaving just one strain (WPV1) which is geographically constrained.

A question for those well-versed in the biomedical and ecological sciences: Could our attempts to eliminate viruses be a bad thing? I wonder about the ripple effects of ecosystems on each other. We do not brag about extinction of any other natural entity.

And, when eradication does not work, when we only hinder the growth of a virus, does the attempt to prevent infection (or any form of "mostly cleaned up" coverage) encourage the development of superviruses by selecting for the strongest in the set, in the same way bacteria do? If not, why?

gburt··on The Sketchy World of Fake Bike Gear
"Surprisingly, Tombragel’s fake fork passed impact testing—but barely. If this had been a real Specialized, it would be re-engineered to pass by a higher margin."

This line is when I decided this was probably a PR placement. I guess that maybe took me a bit longer than most, but it just feels so stilted.

"We passed our test, but we'd want to pass more." - then why did it 'pass'?

gburt··on Soylent Shipping Delay Root Cause Analysis
How could you possibly know this? Where is this comprehensive prescient list of "all the nutrients required for an adult human"?
gburt··on The student who discovered water on Mars says it was a 'lucky accident'
I don't disagree with anything you said. I think the curse of knowledge and Dunning-Kruger describe the same effect (one in a semi-quantitative way and one in a philosophical way).
gburt··on The student who discovered water on Mars says it was a 'lucky accident'
This is more/less the Dunning-Kruger [1] effect in action. From the student's perspective, all of his knowledge is "obvious common sense that everyone has" (this is his error about others).

[1] https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect

gburt··on Breeding the Nutrition Out of Our Food (2013)
How reliable/discrete is this process? Can I select particular nutrients and decide I want a plant to have them? Can you tell us more/provide some interesting references?
gburt··on More Than One Third of Tutanota Emails Are End-To-End Encrypted
The network effect changes (very expensive) which limit adoption of proof of work also limit adoption of encryption. Assuming encryption is adopted, the vendor(s) involved could simultaneously demand using Hashcash or another email proof of work system.

One should seriously estimate and discuss the environmental implications of such a change before working towards it though, I think.

https://en.wikipedia.org/wiki/Hashcash

gburt··on Linux BBQ
I don't understand. Is there a Linux distro out there with too many combinations?
gburt··on Why drivers in China intentionally kill the pedestrians they hit
I'm not the OP, but some data here: http://freakonomics.com/2014/05/01/the-perfect-crime-a-new-f...
gburt··on Why drivers in China intentionally kill the pedestrians they hit
Not if you set the penalty for killing even higher. The problem here is the order of the size of the penalties, not the penalty itself.
gburt··on Why drivers in China intentionally kill the pedestrians they hit
The language "accident" (and the associated story) implies that one cannot control whether they injure someone with their 2000lb death machines. That is not true.
gburt··on Understanding ETF “Flash Crashes”
I don't see that as a problem. The circuit breaker size is 10% (5% in some markets, I believe) for 5 minutes. It would take less than 2 hours to completely obliterate a single ticker, and each market participant gets the "same" amount of play time as before.

It might slightly favor HFTs if things do change across those 5 minute downtimes. Otherwise, your orders should be just as valid before the freeze as after, no?

gburt··on For Evolving Brains, a ‘Paleo’ Diet Full of Carbs
In a space as high-dimensional and full of incorrect intuitions as nutrition science, credentialism is perhaps a worthy signal.
gburt··on Investors Find Ways to Indirectly Profit from Valuable Startups
Indeed. To me, this article seemed like a PR piece designed to increase HomeAway's stock price.
gburt··on Court rules drug dog barely more accurate than a coin flip is good enough
Sorry - I edited my post multiple times after posting it to clean it up and increase quality. Some of your response (especially the first paragraph) is now confusing. That was my fault.

I agree with you mostly - especially on the discussion of what "probable cause" should be defined as and what constitutes a "random" search, as just an example of some of the ethical issues of interest here. That is a debate to have and even one I think this community is well equipped to discuss, but it was not something I was trying to comment on.

I was only making the point that it wasn't fair to jump on the 93% as though it was representative of the problem. I am not taking any ethical or defensive position of drug dogs here.

gburt··on Court rules drug dog barely more accurate than a coin flip is good enough
I have no opinion on the legal and ethical issues, I was commenting only on the classifier. I don't think your first question is appropriately described as "the real question" though, as one of the coolest results of machine learning is that there are techniques for combining multiple bad classifiers [0] in to one really good one.

Even more so, I think that is actually what is happening here. An officer is acting as a tier 1 filter, and then the drug dog is acting as a tier 2 filter, further reducing the number of false positives, which in a "legal and ethical" sense, we generally feel are worse than false negatives.

[0] https://en.wikipedia.org/wiki/Ensemble_learning

gburt··on Court rules drug dog barely more accurate than a coin flip is good enough
This isn't being fair. I don't really think the post title is fair either. That 93% isn't comparable to the 50% you're probably thinking for the coin flip, or even the 0/100% mentioned in your edit.

---

93% alert rate means:

93% = false positive + true positive

7% = true negative + false negative

---

We're given "His success rate was much lower, at 59 percent." Assuming that means that 59% are true positives, we know the false positive rate is 34%.

Without data on how many passengers are carrying drugs, we can't say anything about the rest of that result. A plausible world where the drug dog may be valuable is when he unquestionably rules out that 7% of the population that he doesn't alert on, as in false negative = 0.

I believe the reason this number is so high (59%) is that the officer has already used his suspicion to determine filter out most people. This makes the value of the "ruling out" plausibility even more valuable: the drug dog reduces searches compared to trusting the cop.

Ideally, they'd provide the receiver operating characteristics curve [0] or the confusion matrix [1].

[0] https://en.wikipedia.org/wiki/Receiver_operating_characteris... [1] https://en.wikipedia.org/wiki/Confusion_matrix

gburt··on In search of the perfect URL validation regex
> http://3628126748

This shouldn't fail. It resolves to an IP owned by The Coca Cola Corp (not an internal IP)

gburt··on Training a Neural Net to Generate CSS
My point was that this is boring. A fitness function that evaluated some notion of how "pretty" a page was would be cooler than being able to regenerate a screenshot's CSS (in a likely very complex form).
gburt··on Training a Neural Net to Generate CSS
The "reward signal" aka objective function seems to be the challenging part here. In the parent post's suggestion, all you'd end up with is a neural network that could /maybe/ reproduce a picture (assuming CSS was capable and the network has the approximately learnable properties necessary). It'd be more interesting to have some "quality" measure that actually meant something to evaluate outputs.
gburt··on Index Funds May Work a Little Too Well
I'm not the OP or the grandparent post, but I wonder what happens if a significant portion of the market starts indexing? My gut is that there are strong incentives against that: as more people start indexing, the gains from not indexing become more diverse.
gburt··on How the Witcher 3’s economy was saved by polynomial least squares
This happens in the real world too. Imagine travelling to a poor country as someone easily identified as wealthy, for instance. It happens on a microscale within a given country too, if anyone identifies your wealth, you'll find prices creep upwards, especially for work of a contracting variety.

To really confirm this, try bringing two watches to a watchmaker at different times, one a $5,000 watch and one a $500 watch. Ask what a service costs. The gap tends to exist even if the movements are the same!

gburt··on How the Witcher 3’s economy was saved by polynomial least squares
Saying "polynomial least squares" to refer to a class of model fitting techniques that...

1) basically dominates the entire space of statistical analysis

2) was almost entirely irrelevant to the article

...struck me as pretty artificial to me too.

← PreviousPage 4 of 14Next →