Yeah, the obvious takeaway is compassion, but this experiment didn't have any control condition. If it were me, I'd hesitate, sure, but it could just be because I was trying to process what is going on (whether this is part of the experiment, whether the robot is malfunctioning, etc.). To truly isolate "compassionate reponse", it would have been helpful to have a condition where the robot behaved in an unexpected but less directed way, like flashing lights or emitting odd sounds. I'd bet there would be a lot of hesitation in those scenarios, too...

OK, these are cool images. However, I do think this is a bit biased toward the marketing-friendly description on the product page (marsbioimaging.com).

First, calling this a "color" image is not correct. Just as with radar, sonar, MRI, or anything else that captures non-visible light, what we are seeing is "false color". The distinction is that someone mapped signal values to colors, and that implies that those nice pics probably had some human input to make appropriate color choices (and not blue flesh, red bones, or whatever).

Second, this is not the first spectral CT, as the article seems to imply. Check out for example https://www.itnonline.com/article/spectral-imaging-brings-new-light-ct for a summary of what commercial offerings were available 3 years ago.

Don't get me wrong, it sounds like great stuff -- but there seems to be significant hype here, too.

I often don't click through a link, but instead prefer to copy it, paste it, and if needed modify it. It seems anyone doing some basic "link hygiene" like this might not get counted?

This is from a "Nature Partner Journal" (see http://www.nature.com/partnerpublishing/journals/), which is a line of open access content under the Nature brand, not Nature itself.

Normally when I read an article like this, I go to the Methods first to figure how much skepticism I'm going to have when I then check out the Results. There isn't a Methods section in this paper! I'm consequently not terribly impressed by the quality of this journal.

According to Stanford's Copyright and Fair Use summary (http://fairuse.stanford.edu/overview/fair-use/cases/) that Twin Peaks case might not be the precedent they say it is.

Not a fair use. A company published a book entitled Welcome to Twin Peaks: A Complete Guide to Who’s Who and What’s What, containing direct quotations and paraphrases from the television show Twin Peaks, as well as detailed descriptions of plots, characters, and setting. Important factors: The amount of the material taken was substantial and the publication adversely affected the potential market for authorized books about the program. (Twin Peaks v. Publications Int’l, Ltd., 996 F.2d 1366 (2d Cir. 1993).)

(emphasis added)

The submitter used the word "poorest", which seems chosen rather... poorly. The SFGate article uses the somewhat less extreme term "low income", but toward the end it is also more specific about the criteria: "To qualify, applicants must live in a neighborhood designated as disadvantaged by the state. They must own their homes and make no more than 80 percent of their community’s median household income." The provider, GRID Alternatives, promises "to make renewable energy technology and job training accessible to underserved communities", which seems more in line with what is actually going on.

So, one view of this is that this is a program to direct cap-and-trade money (generally collected to be used specifically for environmentally beneficial projects) into areas of the state that wouldn't get it otherwise. It uses donated equipment and labor as well as the C&T funding, so it's not at all tax funded. Besides helping recipients in the targeted areas get cheaper power, it is possibly reducing overall electricity demand in a green way (though this is debatable, given the limits of solar power as a baseload source).

that 3/4 of people would pick up a penny from the ground? I would pick up a piece of trash from the ground, but that doesn't mean I want trash.

It could be described as a variant of the Traveling Salesman problem, where each node is a mower-sized swath of grass and your object is to visit very node, returning to the starting one...

So, was she impressed?

> It really bothers you? How? Please tell me how I've ruined your shopping experience.

Many of these are obviously jokes, fine. I think what the OP was getting at is that there are nefarious purposes to which this can be put. Imagine if you want to "attack" a competitor, and make up a bunch of poor reviews (there have been court cases about this kind of thing, in this case about libel). Or, you could just boost your own product, or try to game the Amazon recommendation system to get your product recommended based on the fact that you "like" many popular items in a segment, plus this product.

Despite research on the topic, it's not going away anytime soon, just because it's obviously pretty hard to figure out when an "attack" on the system is happening (though there are some clues that have been used in automated detection, e.g. lots of similar rating events in a row, accounts used only to rate a few products, etc.).

They should be sure to ask for the digital negatives. Without those, I'm sure nobody will be able to disseminate these documents any further!

Compression is just the discarding of irrelevant or less relevant information. With images or video, that means keeping the perceptually meaningful content and discarding the rest. An improvement might come about if the encoder was removing irrelevant variations (noise), or smoothing out unnecessary details away from perceptually salient objects (making them easier to see).

It's pretty hard to make an image encoder that maintains the important perceptual qualities of every possible image, so IF their encoder is good, maybe they just didn't test it on the whole range of stuff they eventually used it on.

I sort of agree. I think it's "legendary customer satisfaction" that TFA is thinking of (that's where Apple has always lead, anyway). If you have a Mac you may have the same lowish rate of problems (many of which are possibly component problems and not much to do with Apple), but statistically speaking you're more satisfied overall, which is probably driven by the large fraction *without* problems.