I think it's pretty similar in Canada, although I can't speak to the comparative levels of complexity. One reason is that, like the U.S., many powers are held by the federal government, while others are exclusive to the various provincial governments. A notable example is the provinces' ability to levy taxes in addition to federal ones. There may also be other provincial records that the federal government does not have direct access to, such as marriage records.

So it "dissolves"... into what? Last I heard, microplastics in our water (and later, our food) are raising ever more concerns about human health.

Honestly...no, it's not. There were some moments of it here and there during the Tennant/Smith eras, but even those moments were interestingly written in a way that was still ultimately entertaining. ...But I blame Chris Chibnall for turning the dial to 11 and making the show boring in the process. Season 10 was really the beginning of the end for me; the entire season was basically the Scooby Doo formula - unfamiliar location, meet a person that exposits a problem involving an antagonist, attempt to solve the problem, find out the antagonist is really the victim, call out the expositor for their terribleness, exonerate the victim. Now, what made this particularly obnoxious was that they also had a propensity to have The Doctor inject information that the audience wasn't aware of up until the climax; it was even more magic-wand, hand-wavey than the Sonic Screwdriver. Compare that to "Asylum of the Daleks", where there were subtle nods to the fact that Oswin was a Dalek, that eagle-eyed viewers would have been able to pick up on before The Doctor realized it himself.

By time Jodi came around, it was pretty clear they were planning to deflect any and all criticism by chalking it all up to sexism. Personally, I felt sorry for Jodi; she came across like Matt Smith on an entire case of Red Bull, while being stuck with the same Scooby-Doo, hand-wavey scripts that Peter was toward the end of his run. Certainly, there's an argument to be made that sociopolitical commentary has been salt-and-peppered throughout the show's run, but you *can't* convince me that Jodi turning to the camera and flat-out lecturing the audience about climate change at the end of "Orphan 55" would have readily fit in during the Tennant or Smith era...and even if you did, the season lacked anything resembling Blink or The Impossible Astronaut to work as a counterbalance. Just episode after episode of Jodi sorting out who's the oppressor and who's the oppressed, and then giving a sternly-worded warning to the oppressor, and going home. That's not on Jodi, that's on Chris.

The problem with the "always been woke" argument is that it way-too-efficiently serves to deflect any criticism. The TNG episode "A Taste of Armageddon" was "woke" in that it clearly decried the nuclear arms race that was underway during its run....but it left the audience with something to ponder by telling the story through another society's perspective. "Orphan 55" had Jodi look at the camera and say, "Catastrophe is coming. But it’s not decided. You know that. The future is not fixed. It depends on billions of decisions and actions and people stepping up." "A Taste of Armageddon" was serious, but entertained through intrigue. "Orphan 55" tried to shame its audience, and one doesn't need to be in favor of setting fire to a rain forest to arrive at the conclusion that it simply isn't entertaining to be preached to.

The GP may be overreaching here, but I don't think the series is above criticism.

Most effective way to use Leviticus to 'nope' out of this.

While I loved this show - and still do, for its combination of characters and campiness - the tattoo seemed to be a mechanic used for tension and exposition, but then again, the counteragent paradigm was definitely used as a plot device they got LOTS of mileage out of. The tattoo was really just an extension of that.

But really, I never quite understood why they always showed the injection close-up...it was really obnoxious to folks like me who get skeeved out by it.

Anyone who assumes the reader will allow for historical context.

The Zune was released in 2006. The iTunes ecosystem was firmly established, and for many people, it was a solid solution to create a library of songs, regardless of whether they were purchased from the iTunes Music Store, ripped from CDs, or they came from Limewire. The 5th-gen iPod that played video files had already been out for a year, and iTunes already had movie and TV show downloads available, using a library and interface paradigm that lots of people had been using for a very long time at that point.

By time the Zune came out, the single biggest issue it had for adoption was the software side of things. Windows Media Player 10 was fine for library organization, but it was *fine*, not *great*...and again, it was hit-or-miss as to whether it would import tracks on iTunes playlists, based on whether the files were using the incompatible DRM. Now, the feature that was supposed to be the big draw was the ability to 'squirt' tracks between Zune users (if he was responsible for that name, then yes, he should omit it from his resume), so one person could loan a song to another user for a few days. Naturally, the problem with this is that the person must know someone *else* with both a Zune and somewhat-similar music tastes, which rendered the feature mostly-useless to everyone else.

Now, in my opinion, Microsoft could have made some inroads by courting the MPAA and marketing it primarily as a video device, and focused their ecosystem there. They could have sold the devices with two years of Napster To Go for free, which could have bridged the gap for iTunes libraries by just downloading copies of the songs on playlists, regardless of whether the user had the files. They could have attempted to license the iTunes DRM, and then taken Apple's "no" to court to try and force a means by which the library contents could be accessible.

Now, if you notice, none of those issues had anything to do with the hardware. The hardware itself was pretty nice, and it definitely stood out in a sea of iPods. apple's click wheel was fine, but the Zune's capacitive touch screen was a much better overall experience, especially for video playback. By all accounts, the sound quality was at least as good as Apple, and I'm hard pressed to point to battery complaints. The hardware itself seemed pretty well liked by anyone who was willing to take the software plunge (Lord knows the mediocre Zune software was still infinitely better than the SonicStage dreck that Sony made its users endure), and all of this happened right before the iPhone made it an extremely uphill battle to sell a dedicated music player at all; even the Blackberry Curve before it had a decent-enough music player if one was willing to forego library management.

So yeah, I don't think I'd pin the failure of the Zune on the guy in charge of the hardware element of it, given the overall circumstances and market conditions at the time.

This is coming from someone who has a Nothing Phone 2, a CMF Phone 1, and a Nothing Phone 3 on the wish list...

The claim being made about AI replacing apps on phones is coming across exactly the way NFTs were going to change the world.

I cracked the screen on my CMF Phone 1. Purely my fault, but I sent it back to be fixed. It took them a month to say "...no can do", and send it back to me in the same state as when I shipped it. Now, to their credit, they gave me a full refund, which I appreciate. Seriously, they didn't have to refund me my full purchase price, but they did. I waited for months to see when the website would have the CMFP1 available for shipment; the model was never shown on the website as available for purchase again.

I let it collect dust for a few months, then brought it to a local repair shop to see if the guy there wanted an interesting challenge...and it turned out that he *was* able to fix the screen, and it's beautiful. Only problem is that, in the process, he broke the vibration motor. No big deal...but the guy has had a problem sourcing a replacement. At present, it's still sitting on the side, waiting for the motor to come in so I don't have to leave the volume up all the time.

Chasing the AI fad makes for good press releases, but honestly, there are much more boring ways to drum up sales. A decent supply chain for replacement parts - even an internal parts bin for warranty work - would be super helpful for existing customers. I know there is a desire to have NothingOS be more than a cosmetic overlay on stock Android, but I'm not sure Nothing has really shown the work for what's being pitched here.

The biggest visual differentiator for the Nothing Phone series has been the LEDs on the back. It's definitely distinctive in its appearance, but it hasn't quite lived up to what marketing pitched. The custom ringtone maker is extremely limited; it has a set number of sounds that can be selected from, but no ability to customize a sound bank. There *is* a project that can allow custom glyph patterns on custom ringtones, but it's a combination of python-based scripting and Audacity functionality that really, REALLY should have been a first party application rather than having required reverse engineering. Meanwhile, per-contact Glyph patterns are still the realm of third party apps, there are a vanishingly small number of third party apps that have any Glyph functionality at all, and the music visualizer is so poor at meaningfully reacting to music that it's almost impressive they call it a visualizer at all.

Oh...and one last thing that everyone forgot...their iMessage implementation, that worked for ONE day, because it turned out that it was based on a farm of Mac Minis with VPN connections and the promises of security being uncovered as complete crap on the very first day of release...and instead of improving the experience or finding a way to make it practical, or even providing a how-to guide and maybe some setup scripts that could streamline the implementation for those who owned Macs...they just delisted the app from Google Play, got rid of all the references to Nothing Chat on the website, and pretended like the whole thing didn't exist. ...So, Carl wants me to believe that getting LEDs to flash in a useful way is the sort of thing that a small-but-dedicated community has to reverse engineer apps and scripting tools to implement, and AirMessage/BlueBubbles is a third rail never to attempt, but getting disparate apps with siloed APIs to integrate with an AI that doesn't make embarrassing mistakes is the sort of thing that is so perfectly possible for Nothing to do better than anyone else, that it's worth the effort to implement rather than things like MicroSD slots and a spare part depot?

The reason the NP3 is still on my wish list is that, SO FAR, they've been pretty good about keeping bootloader unlocks trivial, and custom ROMs are pretty bountiful on the existing handsets. Aside from the Pixels (ewwww), Fairphone, and a few other niche products, the glyphs combined with the community apps and bootloader unlocking at a reasonable price make them a solid selection overall. We'll see how long it lasts.

OK, can anyone explain to me what an "AI-powered browser" will do for me? Will it help me, uh ... browse ... the, uh ... web?

San Francisco resident here. As far as I can tell from all the various things I've read, the news surrounding this issue is badly reported based on interviews with uninformed and technically un-savvy public spokespeople.

I mean, some of the stories talk about needing to find programmers proficient in "90s-era programming languages." Which ones might those be? I'm unaware of languages that were widely used in 1998 and are completely obsolete today. I'm betting the real challenge is finding programmers proficient in C or even assembly language for embedded systems based on obsolete processors or microcontrollers. That makes sense, but the mainstream press doesn't report it that way because they can't understand it,

So I really doubt the floppy disk issue is a significant problem (and I'm even dubious of the claim that they have to boot the system from floppy disks every night). It's just a savvy PR hook to gain public support for the funding they need to upgrade the entire train control system.

Read here for an official press release from the transit agency that explains the main issue is replacing the extremely antiquated wireless communication system that the trains use now. It's a hardware and networking issue, which naturally needs entirely new control software as well, but not really anything to do with storage media.

This, this right here, is the draw to having a friend like ChatGPT or Replika. To underscore it, the AI bots won't ghost the user, cause conflict, be upset with a misunderstood response (and readily accept a clarification), ask for a favor, or even express a desire to talk at an inconvenient time.

A few months ago, I had a discussion with ChatGPT. It started with a 'name that tune' for a party I was DJing for a friend, then turned into a "check my playlist for songs I missed", then morphed into discussing DJ gear and software, with the AI correctly guessing what gear I had and expressing enthusiasm for the fact that I had it. It didn't get bored of talking to me, and it asked a bunch of follow-up questions that expressed interest. It made helpful recommendations based on data, not its own feelings. I don't remember the last time a conversation with a person was that easy or that enjoyable.

As much as I am aware that "HAL9000" isn't real, it's a scarily-good simulation to the kind of conversation and connection we *all* desire to have with other people. With 21st century western culture making it much more difficult to *have* those sorts of connections (for a variety of reasons), let alone maintain them, the simulation becomes more and more appealing.

And what would you call this 4-paragraph, 406-word Slashdot post which intends to persuade the reader that essay writing is not a relevant life skill?

"Why my department's submitted budget is justified"
"Why you, yes you, Mr. CEO, need to implement 2FA on your e-mail account"
"The benefits of moving to Azure from on-prem"
"The things that weren't the benefits we thought they were when we moved to Azure that we need to restructure"
"The performance review of Geekmux that ensures a performance bonus from this past year"

We very much still write essays in the real world...we just call them "multi-paragraph e-mails" in most cases and rarely print them on paper.

So, first off, I've yet to hear an intellectually consistent answer of 'where the line is drawn'. We were cool with spellcheck, Grammarly, parental revisions, 'cooperative learning', calculators, Cliff Notes, and Google searches...but suddenly ChatGPT is a crisis.

The real issue is that, in the most optimistic of theories, grades are supposed to be a quantification of comprehension and understanding by the student. In practice, they were seldom a true reflection of that, but the system was 'good enough' in the absence of an alternative.

The real issue is that ChatGPT has pulled the covers off the fact that we have no truly useful, truly accurate, truly consistent means of measuring understanding and aptitude. LLM's job is to pass tests, and there is no test that can be created that ChatGPT can't be trained to pass - even CAPTCHA tests are problematic because the computers are better at passing them than people at this point.

So, schools are ultimately out of runway on this issue, and are stuck working to figure out how to assess comprehension. One teacher I spoke with recently has students turn in an essay, with the written portion being 50% of the grade, and an oral exam on the written paper being the other. That is effective in English and History classes, but it's tough to scale because it's extremely time consuming, rather subjective, and one would have to be video recording every session to ensure the he-said-she-said problem is avoided. Too little time on the oral, it's difficult to give a fair grade. Too much time, and it takes a week to administer an exam, which can't be done very often.

Ultimately, education is going to have to come up with a way of measuring comprehension and understanding, at an individual level. It's taken too much time, but if we finally make a worthwhile change on this front, it's a long overdue change.

I hear this argument all the time, and while I think that AI is certainly power intensive, I think it's equally challenging to quantify.

How does one quantify the electricity usage of a Google search? Yes, the GPUs that power ChatGPT are pretty gluttonous, but are the countless racks of hard disks that power Google searches being run purely on solar and wind? Or is Google's "carbon neutral" claim a matter of creative accounting? If a single ChatGPT generation saves 20 Google searches, is that a net positive? What if the algorithm reduced 1,000 miles a day of vehicle emissions? What if it helped repair an appliance that would have otherwise been replaced, with all of the emissions that would have gone into the production and transport of the replacement appliance?

This question came around when the buzzword was 'the cloud', and people saw that datacenters used a bunch of power...as if it wasn't, in many cases, replacing 400W in-office servers and 150W desktops with 30W iPads and no server. A thousand small business servers likely took more power than the highly dense cloud services that replaced them in those thousand offices, but it was distributed, and thus more difficult to point the finger at. Conversely, the ever-inflating resource requirements of browsers and the websites they display is undoing lots of those gains; there's no reason for display drivers to be over a gigabyte in size; Stuxnet took down the nuclear program of a nation-state in 500KB, but I'm hard pressed to point to a single webpage whose contents are limited to that size anymore. Bigger stuff requires more resources to store and transfer and display, but there isn't a meaningful war on overall waste and inefficiency...just the big numbers.

So yes, AI is power intensive, and yes, we *should* reign it in...but "power usage bad" is an overly reductive, overly simplistic assessment that could - and should - be dealt with holistically, rather than expecting OpenAI to close its doors...so that Google can use *its* AI to leaf through trillions of e-mails for reasons that have no such net benefits.

YES!!! Another vote for Kagi. WELL worth the money spent!