> Another thing is that trying to take away creative jobs from humans and giving it to robots while the menial ones (like data entry, manual labor, household work, etc.) are still being done by humans is an insult to humanity itself.
I don't think this is going to happen anytime soon. LLMs may be able to somewhat compete with people who are average at these creative tasks, but those are not the people whose work most people consume. We listen to music and read books by people who are very good at creating music/writing, and I don't see LLMs competing with those people. And even if they were able to compete, they'd still miss the "human" element - eg., I'm probably not going to a concert if there are no humans performing.
AI will expand to be able to do more boring, menial tasks. Writers and musicians (and other creative people) will probably be amongst the last who loose their jobs to AI.
I think the problem here is a societal one. One born of capitalism, that does not value art in itself but only the value people ascribe to it.
And the problem I see is less that it replaces good artists, but that it takes away menial work that artists might use to support their more artistic endeavors.
I think this is a real problem worthy of much discussion, but I also think this is not primarily an AI topic but more a general one about what kind of human expression our societies value and want to encourage. You can't stop progress here, though we could decide to not like it and do something else.
I don't think a simulation needs to simulate anything real. Eg. if I have a program that runs Conway's game of life, that would be also be a simulation, but there's nothing "real" it references against.
In other words, the universe simulating ours might work completely differently, and some beings there are just testing how different laws of physics (i.e. ours) work out.
some beings there are just testing how different laws of physics (i.e. ours) work out
Or there is a practical use like energy extraction. Or they live (if that even applies) in a universe that generates what our closest concepts of are e.g. time and energy, and they have to spend them massively in simulations to survive. Or their laws of nature tend to overcommunicate (dilute) them into a single being, so they simulate everything we see for islands of “themselves” to remain distinct. They split themselves endlessly with perception because that’s the trick in their situation.
One of my nightmares is to wake up from this and realize you woken up from a mild depression into a horror of a true existence, and this isn’t even a final level. Makes sense, since if god is infinite, why shouldn’t he be infinitely suffering. Why would he (you) create all this if not to escape from null-reality.
Afaik what makes fusion hard is maintaining the conditions that allow fusion to occur. Because of this, it’s quite safe - if anything goes wrong, it’ll just stop working.
Yes they do exist, link elsewhere in this thread, Helion have a 7th generation nuclear fusion reactor fusing once every ten seconds or so. https://www.youtube.com/watch?v=_bDXXWQxK38
(It's not surplus of power, but it is repeatable fusion actually happening).
The page turn animation was truly excellent. I have seen a few other apps/sites try to copy it, some of them did it pretty well. But one detail I’ve never seen anyone else do: the text of the curled page is “distorted” in 3D, as one would expect of a real page. The closer the letters are to the part of the page that’s orthogonal to the screen, the more the are squished.
It’s an effect that’s quite complicated to do. You need to put the page on a 3D cone and render that. I have quite a bit of experience with UI kit animations, but I don’t know how I would do that one.
iPhone iOS design used to be excellent.
Since Apple switched to flat design, it was very clear that they were heading toward this path. I wrote about it in 2016:
“ I predict that the technological disparity will increase dramatically. Most of the efforts that Steve Jobs put to “push the human race forward,” by making tech products easy-to-use to everyone, will be wasted.”
At some point page turning as in physical book becomes an anachronism like the floppy disk icon on save buttons. Do you expect your browser to flip pages upon navigation?
I don't expect this will happen. There's a good reason why floppy disks are no longer used, but I would be surprised if physical books ever go away.
Personally, as much as I like the convenience of digital books, there's nothing quite like the reading experience that comes with a printed physical book.
Sure, but how much closer to the physical book reading experience does a page flip animation really get us?
In my experience, what makes a digital reading app pleasant to use are other things: Responsiveness, choice of default fonts, whether or not it allows the publisher to apply ridiculous style overrides (e.g. forcing the background to be light grey even in night mode), not losing track of pagination at chapter boundaries (so that flipping back across a boundary, paragraphs don't mysteriously shift up or down half a page)...
I'll take a reading app that does these things well, but only uses a simple slide animation over one that imitates a perfect page flip, but otherwise botches the concept of a "page", any day.
Glowing in the dark is as much a bug as a feature. Lit screens work best in a dark room, where a reading light is more pleasant when you pause reading to do something else.
I spend an inordinate amount of time either turning lights on an off or trying to use the not quite sufficient tablet screen to do stuff.
An e-ink reader does it better because the front lighting is less glaring and the overall image is more stable. It’s hard to get a phone screen to the right brightness level for reading in the dark, but an ereader will have a very dim front light that you only notice when the room is completely dark.
Kindle Paperwhite 5th gen has both amber and white LEDs and its color temperature can be adjusted. I don't know exactly what light temperature range it has but it feels very "warm" on maximum setting.
There is something about owning a book and having your own hard library. I love to do almost everything on the phone; news, tv, videos, Ect.
Reading from a book however is a better experience than a lit up screen. Audiobooks However have a place as less and less people have the attention span to physically hold something and read it.
User Interface that is easy to imagine as a physical object is easier to figure out. It doesn't matter if the user is familiar with the physical object being emulated on the screen. Because laws of physics are still applicable even if you haven't seen such a physical object in real life.
The human mind is always building models in our minds, based on observations of how things look and work, and then using those models to predict behaviors of things that we are yet to explore. For this to work well, appearance has to imply behavior. Instead of starting with a clean slate each time it would be helpful if we can leverage the mental models we have already built of the physical world. This is where skeuomorphism is helpful.
I don’t have any data to back it up but I doubt that. Schools are still mostly using paper text books. And I reckon school libraries are still the main source of books for kids (who typically don’t have money to buy books). Family members who want to buy you a book as a gift, will do that on paper as gifting ebooks is trickier and less personal.
To be honest, I would bet that the if younger demographics are reading less paper books it’s because they’re reading less generally rather than switching to digital.
True. But piracy is a big issue, and it exclusively happens digitally. I wonder what the numbers would look like if you factored in pirated ebooks/etextbooks.
I don't know if it's nostalgia or what but the flat design seems charmless compared to the Forstall-era skeuomorphic design. And the vanilla sameness and uniformity makes it harder to distinguish things. It reminds me a bit of shelving books by size and color rather than by content.
I recently noticed a bunch of electric street lamps that are clearly modeled on gaslamps - they appear to have little chimney vents as well as wide pipes for the nonexistent gas, and they seem designed to shield the LED/CFL/etc. fixture from wind and rain. Mostly non-functional design elements, and I've only seen a few dozen actual gas streetlamps in cities, but I still find them charming.
>It’s an effect that’s quite complicated to do. You need to put the page on a 3D cone and render that. I have quite a bit of experience with UI kit animations, but I don’t know how I would do that one.
It's not possible with public UIKit / CoreAnimation APIs. Those only support homogenous linear transformations (i.e. 4x4 matrix). You may try using the private CAMeshTransform API to achieve such an effect: https://ciechanow.ski/mesh-transforms/
You can actually do it in UIKit by chopping it up into strips and using affine transforms, I've seen it done. Not that I think this is UIKit, it's probably a very simple OpenGL shader.
But did it alternate between the left and right pages? The swipe from left page to right page shouldn't have been a page-turn; it should be a horizontal scroll across the book's binding. The swipe after THAT should be a page-turn.
maybe i'm just unsophisticated but what exactly is the point of it? all this work for basically nothing. and we wonder why our apps are bloated and need ridiculously overpowered hw.
There is something about reading on a screen which just isn't quite as "good" as reading on paper. I don't know what it is.
Is it the reflectance/emission? Perhaps, Kindles are better than iPads; is it the resolution? Perhaps, retina iPads are better than pre/non-retina tablets; is it the tactile sensation? Perhaps, I find matt paper better than the gloss of many magazines, and the new Paperwhite is half way between on that score.
Perhaps I'm just remembering good times from my childhood, and skeuomorphisms are a way to catch that.
But no, a 3D animation like this is not the reason why apps are bloated. Other similar animations were smooth on a 450 MHz G3.
> There is something about reading on a screen which just isn't quite as "good" as reading on paper. I don't know what it is.
For me it’s the exact opposite. I read a lot, mostly on my Kindle Oasis (139 titles this year so far according to Kindle Insights) and on the very rare occasion I read a paper book I’m reminded how annoying reading paper books is.
There are several issues with reading paper books:
First, the physical format, long books are thick and unwieldy. There is no comfortable way to read in bed. You either read laying on your back, holding the book above your face, which is uncomfortable to hold and tires your arms. When laying on your side the fact that books fold in the middle is super annoying, if you open the book at a 90° angle you can only really read one page and you have to turn yourself after every page. Holding it open fully also isn’t comfortable.
Next, there is the light issue. Paper only reflects light, meaning you always need an external light source. It’s much easier for me to immerse myself in a story reading in a dark room. Another issue with external light is that you have to orient yourself relative to the light source. Again, when reading in bed this is a problem if your light source is on your nightstand. If you turn to a different side you are lying in your own shadow.
Last, there is the problem of logistics. As I said I go through a lot of books. If I had to buy these physically I would have run out of storage space years ago. Books would be piling up all over my apartment. Getting my hands on them in the first place would also be a problem. I can browse books online and find something I’m in the mood for right now and be reading it in 30 seconds instead if waiting days for delivery. I can binge through a series in days instead of weeks.
No, I really don’t want to go back to dead tree books and I can’t believe people put up with the inconvenience when there is no longer a need to.
i mostly agree... i think a lot of people haven't tried the more recent higher-end e-readers... i'd say both the oasis and the kobo forma are getting quite close to strict improvements on the trade paperback... i still love a good, well-worn <400 page mass-market paperback sometimes though, truly the greatest height of the dead-tree format. e-readers haven't seen dramatic technical improvement, but response times have gotten a lot faster, and manufacturers have pretty much all finally figured out physical buttons for page turning is the way to go.
last time i read a trade paper, the weight and lighting issues dominated any of the pleasant tactility of paper.
that said, i do miss the random-access characteristics of physical books: being able to have fingers between multiple pages to skip between sections and the ability to quickly visually binary-search for something. these seem solvable but require some master UI work
> Last, there is the problem of logistics. As I said I go through a lot of books. If I had to buy these physically I would have run out of storage space years ago. Books would be piling up all over my apartment.
The other side of that is discoverability. So much of my fondness for reading today comes from the fact that I grew up in a house packed with books, practically a shelf in every room, and as a child it was so easy to just pick a book off the shelf and start reading.
The Kindle Oasis is my preferred way to read now, but if when I was a child my dad had read everything electronically, would I be reading (e)books today, or would I have never read enough casually to develop a love of the medium?
That’s why in 2022 I keep the bookshelves in my house overflowing.
Likewise, I find that browsing through a bookstore or library is still much more enjoyable than trying the same thing with an online shop when I don't have a certain book in mind and am just trying to discover some random bit of reading.
It's because paper doesn't EMIT light; it only reflects it. This simple fact was ignored for the last 30+ years of OS vendors pushing inverse color schemes on us. I see it as a vestige of the "desktop publishing" fad of the late '80s/early '90s, which sought to make the computer screen an analogy for a piece of paper. Or Apple's attempt to look "different."
Now all of a sudden people finally realized that reading dark text off the surface of a glaring light bulb all day is a shitty way to work, and vendors have backpedaled clumsily to offering a hard-coded "dark mode." But we already had an even-better solution: Windows let users set up their own system-wide color scheme, from Windows 3.1 through XP or even Vista. Any properly-constructed application would inherit the system colors for various on-screen elements and guarantee legibility. If you wanted to change the look of all your applications, you had one central place to do it. And if, as a developer, you wanted to guarantee a color scheme, all you had to do was make sure you overrode both foreground and background colors.
But Microsoft actually REMOVED that capability just in time for it to become desirable to more people than ever. Brilliant.
>It's because paper doesn't EMIT light; it only reflects it.
What's the difference? I've seen it repeated over and over and it never made any sense. If your screen is too bright compared to the environment, turn it down. The only practical difference seems to be uniformity - screens are much more uniform. No shadows, no dependency on the angle to the light source etc. Which is... great?
The first things that springs to my mind is that the spectrum of emitted light from an active display is different from that of reflected sunlight/artificial light off of paper made from pulp/linen/cotton.
Active displays can be fatiguing for a variety of reasons, including brightness as you mention, but also the "unnatural" light spectrum.
Also, of course, blue light is believed to affect your circadian rhythm, so can cause disturbances after sundown. Most ebook readers have adjustments for this though.
It's a good question. One thing is that the brightness of the page will always be appropriate for the ambient light, since it is determined by that light.
Beyond that, I'd speculate that screens don't retain as much contrast when you turn their brightness down, compared to physical materials presenting similar "brightness" under ambient light. But really that's just some talking out the ass with no research to back it up.
Screens aren't precise and aren't natural. If you have good enough vision or a bad enough screen, you see that it's a bunch of squares that try to imitate shapes, and even have spaces between them.
Also, the lighting looks fake. Even on good modern screens, there's a billboard feeling I can't ignore. I think that's your reflectance/emission point; it's not light reflecting on an object like literally everything you look at, it's an object blasting light at you trying to make it look real.
Even many low-end devices have better screens nowadays than to see pixels themselves. Also, I don’t get this “natural” point, reading itself is unnatural and not too healthy, not because of light reflection vs emission, but because focusing on one point for a long time, which is the same in both types of reading.
The books underlying required even more work and they’re just a bunch of bytes that accomplish nothing. Some of them are even about something made up entirely and serve no purpose.
This animation was super smooth on even the slowest iPad.
The point of it is: they were going to do a skeuomorphic animation, and they put the effort in to do it right. It's a level of polish that I really appreciate, even though it isn't really necessary for anything.
> even though it isn't really necessary for anything.
It is one of the best, if not the best, way of letting users know what their action will lead to, and of letting them explore the range of available actions.
By behaving similarly to a well-know interface, it simplifies discovery, lowers friction and reduces frustration.
So when smartphones and touch interfaces were new to most people, behaviours such as this one were major, although nearly invisible, product features.
Now? Yes, it mostly isn’t necessary to anything, but it was nice.
I have one app that does it, it doesn't do the corner curl, but it does do the side curl with the reverse text showing on the other side. It uses OpenGL to do it and that is how you can do it.
UI designers seem to bring about changes now, just to show you that they are busy somehow or that there's sorts of progress. The "Second Watershed" as Ivan Illich describes it in "Tools for Conviviality" has reached the Computer User Interface Design as well. How much more can they mutilate it, before they improve it by going back to iOS 1 design paradigm?
It’s a question about ethics, I think it’s implied that legal consequences are not part of the consideration.
But even when including possible jail time in your consideration: if you think that saving two lives at the cost of another is worth it, surely a life saved is worth more than a jail sentence?
I don't think #1 matters, because it runs again once it's in a valid state again. Your test will fail while in an invalid state, but that's to be expected.
As the author explained in another comment, it initially generates the code coverage for all tests. When code is changed, only the tests that cover that code are rerun.
This tool doesn't run all the tests, only the ones affected by the code changes since the last run. It figures out which tests to run by initially running all tests, and storing the code coverage of each test.
It doesn't need to rely on "test change likelihood" - if a code change is outside of the code coverage of the test, it doesn't affect the test.
> If a code change is outside of the code coverage of the test, it doesn't affect the test.
This assertion questionable, the extent it is true it is language specific.
Here is a counterexample in Java:
// foo.java
public class Foo {
public static final int LENGTH = 10;
}
// bar.java
public class Bar {
public int getFooLength() {
return Foo.LENGTH;
}
}
Code coverage tools will not highlight Foo as being touched when Bar.getFooLength is invoked, even though a test asserting that getFooLength() is 10 will fail or succeed depending on Foo.LENGTH. The reason for this is that Foo.LENGTH is inlined in Bar. If you for example wanted to patch a new version of the value, you need to supply Bar.class to affect the change; patching Foo alone would do nothing.
C and C++ likewise do not satisfy this requirement because macros.
You'd want to augment the code analysis with the dependency graph of object/class files (or use it instead of what the OP is using). I'm not sure if it's bullet-proof, but except for runtime dynamic linking, you should get a superset of all affected changes if you just use... whatever your IDE uses to determine which files need to be rebuild when you make a change. Following that graph should give you a superset of all unit tests that need to be run.
Like, e.g. if I change a macro in a header file and press "rebuild" on the project, MSVC (or MSBuild driven by CMake) will figure out that the header was changed, chase down which translation units include it directly or transitively, and rebuild those, then link the output and... chase down everyone else who links to the output and relink them, etc.
I bet you could produce a counterexample that breaks this mechanism (C++ being what it is), but I don't expect to see it in an actual codebase.
I folded his hydrangea design recently, following a video tutorial[1]. It is amazing. The finished model is a beautiful fractal flower. The folding steps aren’t terribly hard to do, but robinhouston‘s description - „magical and startling“ - fits this model very well.
I don't think this is going to happen anytime soon. LLMs may be able to somewhat compete with people who are average at these creative tasks, but those are not the people whose work most people consume. We listen to music and read books by people who are very good at creating music/writing, and I don't see LLMs competing with those people. And even if they were able to compete, they'd still miss the "human" element - eg., I'm probably not going to a concert if there are no humans performing.
AI will expand to be able to do more boring, menial tasks. Writers and musicians (and other creative people) will probably be amongst the last who loose their jobs to AI.