The rear view mirror blocks visibility of what's front and to the right. This makes the car rather unsafe, and it feels the mirror could at least be positioned so that it's slightly higher without obstructing the view.

This is in addition to cars being slightly difficult to see things that are up close.

They're LLM resistant, but not WolframAlpha resistant.

Just before LLMs, there was image recognition phone app that would scan a question, and be able to solve it perfectly. It was designed for calculus-level math, which was much less likely to have word problems.

You know you can just say your a government boot-licker and save us all time.

Or I could point out such claims is used by ones who think the government is one monolithic entity, at the federal, state and municipal levels, along with thinking that the courts will always backup the president and legislature, and also that it somehow spans multiple countries.

It's a little red, white and blue, without the stars and stripes.

major_hellstrom_sees_three_fingers.jpg

But you keep blindly trusting the government,

False, because I pay attention to what's going on.

I expect the US Federal Government to be hazardous for at least 2 years, maybe more. Basically, they claim to have some constitution, but are disregarding its words, and are doing noting to stop problems. By extension, those who caused the government to act in that way (i.e. the voters, especially third-party who dick around, or voters who skipped) are also bad judges of character.

The rest of us will keep working to try to keep our privacy and liberties intact.

To do that, you need to identify the most progressive individuals, and get them elected in a government that you declare is untrustworthy. Then next election, people listen to the "government is untrustworthy" propaganda, and elect the worst possible choice instead to have everything rolled back and made worse.

It's inevitable that you end up with one extreme or another when it comes to moderation

Sounds like one of those Enlightened Centrism complaints. Or a variant - in this case, implying there's only two options, and then guiding people towards the wrong one. Others might call it JAQing off (or "just asking questions" as a means to artifically inflate doubt.)

If something is owned by a democracy-based government, anything that is problematic can be corrected by voting in a different person or party. Likewise, it means you personally should avoid voting for problematic parties, as demonstrated by what happened in the USA.

Do we want to trade corporate surveillance for actual surveillance?

It's a stereotype that governments must do surveillance on every interaction. Even then, a government will immense amounts of data would know not to allow it to lea

But more importantly, that complaint is irrelevant. With social media, it's likely going to be viewable by anyone, which puts the difficulty of surveillance as easy as watching someone's twitter feed. In case of governments, simply tell a private social media to provide access.

Also who is going to use it?

For one, the government can, simply by making it their official social media contact point.

In case the government is using a type of Mastodon/Fediverse node, then users of Mastodon/Fediverse would also be using it as well, simply because another user would boost a statement by a government worker.

AI makes up things 100% of the time.

That's just a feature of Large Language Models.

Game-playing AI instead tries to figure out the "best" solution based on what inputs it has. If it tries making something up (e.g. attempts an illegal move in chess), the game server will catch it and simply reject the move, while other AIs will have already looked at more valid options.

A hallucination is when you don't like the answer.

No, a hallucination is the AI generating something completely false, as demonstrated by yet another lawyer using an AI text generator to invent fake cases and passing them off as real.

The issue with DOS is that the development tools basically ceased development around 1995.

DJGPP lasted until 2000, which is about one year after the source code of Quake was released. IIRC, it could compile the game without difficulty. And... it was an upgrade compared to pirated versions of Borland C++, Qbasic, etc.

but software written for newer tools is usually not compatible with tools from 1995.

Software written for newer tools won't be comparable with MS-DOS, considering that they usually rely on graphics, network, and many other features that expect a modern operating system.

If you have MS-DOS, you need specific libraries, or you do stuff manually. That is, manually call the interrupt table, make hooks, and so on. Also, hope that you used the right functions to unhook the interrupt table in case the game crashes, otherwise you also crash the system and need to reboot.

Brain rot â" the inability to think deeply after too much scrolling on a phone

Brain rot was already occurring without smartphones, and it's known as anti-intellectualism. Mock those smarter than oneself for being nerds, bully them, assault them, and let idiocracy run amok.

The only difference now is that it's commercialized by hyper-stimulating algorithmic social media, and getting rid of any human review or consequences. In fact, the record to speedrun to fascism is held by a Youtuber who simply follows the suggested videos.

In Civilization nobody honors a peace treaty with you for very long.

In Civilization 1, peace treaties only need to last 5 seconds. Simply overpower the opponent, accept their surrender (where they give you all techs and all money), then attack them again. Absolutely no penalty.

In contrast, Civilization 2 requires players to be spotless in order for their word to have any weight. In contrast, AI players can break treaties whenever they want.

I believe the AI browser can hallucinate websites when you need to visit them when they're down.

PhysX is technically middleware, thus there's nothing technically preventing another developer from creating middleware replacement. It might take a while (considering there's multiple versions), but it's technically possible to get something to work (perhaps a PhysX->CUDA mapping.)

But also, it can happen to open source software as well, and I've encoutered two of them.

The BSD Games collection often uses Curses, which is still around, but the latest version available removes the feature of replacing stdscr, and thus plenty of the old games won't even compile without having to scour the source.

Then, Cardinal Quest used Haxe - which changed something in the programming language, and thus can't be compiled without having to make extensive changes in the source code.

But the main issue is that the older games don't get maintained or updated whenever something like this happens.

The training is ineffective on those who could spot a phishing message, and not those who keep falling for scams. Likewise, the ones vulnerable to phishing would still be vulnerable to other forms of scams, such as being asked by a customer to change the shipping address of something already in transit (then doing a chargeback).

What the training does - require skilled users to constantly monitor e-mail for things that they wouldn't normally receive. If the training is poorly designed, then it trips various false positives that increase the work for skilled users.

What can be done is minimize the damage caused by someone getting phished. That is, something similar to 2FA to prevent arbitrary logins (or at least make it obvious), making sure that nobody has full access to everything at once, etc.

Ontario has a rather unique computer, a keyboard, trackball, and various edutainment games. These included Math Maze which had the student enter math questions to defeat opponents, games in the Bartlett Saga where the players were loyalists that had to move into Canada, and a few others.

The software isn't avilable even on abandonware sites, and the system hasn't been emulated. As such, the software is technically gone forever.

I know they were getting Doom running lots of places, but I don't remember a Doom playing AI yet.

There was some basic deathmatch/co-op bots at the time. They could play long enough to kill three monsters, but weren't smart enough to figure out how to win a level beyond flipping random switches, and often given up and just let the human do it.

Best option is the Quake 3 or Unreal Tournament bots.

Then they should design their product so that it still maintains basic functionality.

You may notice this is the same type of demand used for The Crew - that game was rendered inoperable because the multiplayer servers went offline, and there was no reason beyond DRM that it shouldn't be possible to run in single player.

Likewise, the emotional support robot could still have a published API, or at least have enough function to work in offline mode, such as having fallback code similar to an Elisa chatbot.

8. Verifiers and CSPs SHALL NOT prompt subscribers to use knowledge-based authentication (KBA)

So no more requiring users to include the solution for a chess puzzle in their password., nor requiring users to know the location of Google Street View. I'm sure one can live without that.

(e.g., "What was the name of your first pet?")

Oh, that knowledge. That would be slightly harder to shake off, as that's basically the easiest way to get around a forgotten password, and there's still no easy method that keeps unauthorized users out (especially with social engineering tricks, etc.)