Also it doesn’t help that, for example, at U-Haul, employees just use their personal phone to scan a QR code and take pictures of your ID.

I expected the bottom to say something like “instead of cloud-native scalable microservices using AI models that cost an average US household salary per day to run”, but maybe I’ve been in corporate for too long.

As a software developer, this has got to be one of the most original unintended usage scenarios / reasons.

No amount of planning or testing in software will prepare you for ways users will actually use it.

“AI” is a pretty bad blanket term. BirdNet / Merlin is a model trained specifically on bird calls (and some noises) built to recognize bird calls. App and logo use huge general-purpose models to create content of questionable quality.

And yes, a decent birder should never trust just Merlin ID.

This reminds me of “bugoga” from the olden days.

(Circa 2007 on bashorgru, there was a meme about writing “bugoga” into a text file, renaming it to exe and then running it. That made the cursor jump around the console and if you had a connected matrix printer, it would start going crazy - something about writing spaces to all available output in an endless loop)

Someone I know got screwed over when 3G shut down. Their business used an alarm system with a 3G modem (using a cheap data-only SIM card).

The “updated” version requires subscriptions, an app, and goes through manufacturer’s servers, so more expensive, and several more points of failure.

Ironically the free-but-with-ads Tubi is the least bad when it comes to UI. They have categories with exhaustive lists, and even labels for movies highly rated on RottenTomatoes.

Netflix sucked for a while (don’t have it anymore). Apple TV is absolute shit now (subbed for a month to watch a few shows)

Thank you for the warning, that sounds incredibly frustrating. I see Blue Prince recommended alongside Outer Wilds as “knowledge based” puzzle games, but it does not sound fun now.

I also could not get through it and gave up. Some roguelites (Everspace) give you enough upgrades to at give you a better chance of making it to the end (and/or to breeze through the beginning), but weapons, upgrades and blessings in Hades seemed rather marginal, and I am not that good.

Destructible walls. It’s possible to destroy some walls, but that is not in the manual nor hinted at in a classic way (discoloration or cracks). You can complete the game without knowing that. It simply did not occur to me to throw dynamite at blank walls.

Invincibility toggle is nice to have, but even more of a statement of how bad the balance is. (Didn’t need to use it though - found an exploit to instakill the quarry boss :)

Not an avid gamer and play mostly indie.

Tunic. Was recommended for fans of Outer Wilds (fantastic game), but I see no similarities. Tunic’s cover art looks like a Zelda ripoff, world borrows from Hyper Light Drifter, and the game plays like a Zelda + Hyper Light Drifter ripoff, but very poorly tested, unfair, and misleading at times. I get the impression that game designers had zero external opinions and very little playtesting.

Exploration is “bump into every nook and cranny to find a way forward”. Bosses are absurdly difficult, clearly catering to the “git gud” crowd, and do not feel fair (as a counterexample, HLD felt difficult but fair). Puzzles are all backloaded to the latter half of the game, they are mostly fun, but at least one relies on a mechanic that is never hinted at nor used in the game otherwise. The untranslated manual, main gimmick of the game, is misleading at times. Midgame apocalypse / dark world is not that, it’s… Night time (not putting the spoiler because that is just absurdity)

Yes, I finished it, and got the good ending, which only elicited “Um, ok?” from me.

The shape goes into the shape press that presses the shape into pressed shape.

People already have been brainwashed on “how many years of security updates does this smartphone model have?”, and being paranoid about older-than-a-few-years devices, so you’re not far off.

(Not trying to diminish the necessity of software security updates, but at the same time I could not find any Android exploits that do not involve users installing malicious apps, supply chain attacks, or state actors)

What makes this worse is that a lower standard applies to manual searches, allowing the government to conduct extraordinarily invasive electronic device searches without any suspicion of wrongdoing, simply because the border officer chooses to search by hand rather than with a forensic tool.

If those “improperly adjusted” or pointing forward headlights weren’t the same height as my eyes, they would not be a problem. Which is my exact point.

A thought occured to me the other day: there is absolutely no reason that headlights have to be on the upper part of the front grille. Mount them as close to the ground as possible. (Like busses and some cargo trucks) At least it will address a situation where pickup truck / SUV headlights are at eye level for drivers of anything smaller.

Of course! Dark Streets on Lexaloffle

Plays in browser, you can see the code right there too, or LMK if you want a more-commented version (had to remove comments to save space)

Ukrainians would also want to rm rf

I wrote a computer game (including the engine), and matrix / vector math plays a key role in movement and rendering. Definitely needed some refreshers, but the good thing is it is not abstract anymore like it was in school, all math things are directly relevant to in-game objects and mechanics.

First we had Merlin Appholes who would submit misidentified and similar sounds as “sightings”, but Cornell tightened the tolerances on the recognition model.

Now this. But to be fair, even before AI there was a subset of “birders” who loved to exagerrate via ignorance. Cardinal = scarlet tanager (it’s red), american robin = baltimore oriole (duh, orange), red-winged blackbird = Bobolink (it says this meadow has Bobolinks, therefore, any large bird on grass = bobolink), and so on.