It wouldn’t, a simple finite state machine that any intelligent entity could emulate would be enough.

But people have completely deluded themselves into thinking that (what CEOs and marketers call) “AI” is actually intelligent, and this case study shows how preposterous that fantasy actually is.

Flip-flopping in and of itself isn’t bad.

What it is - is a symptom.

A symptom of being an absolute dumbass. Now that is what’s bad.

and those who prevent history from being taught, want to lift themselves up at the expense of others

Does everyone who’s following the old account automatically refollow you when you do that?

It doesn’t port over any old comments/posts, but I’m pretty sure that when anyone @'s you, it’s forwarded to the new account.

IMO it’d still be useful to be able to use an identity you control, like a domain name.

Mastodon already does this

It’s worth pointing out that while ActivityPub doesn’t currently support account migration (although there are proposals in the works for how to do this), Mastodon does have a weak form of support right now.

You can create a new account on another mastodon instance, then you’re able to point your old account to your new account.

Looking back at history, it would lead to more propaganda and more support for going to war.

A population getting attacked only leads to that population wanting to an us vs them mentality and emotional knee-jerk reactions over rational responses.

Because: “The dose makes the poison”.

In other words, any chemical—even water and oxygen—can be toxic if too much is ingested or absorbed into the body. The toxicity of a specific substance depends on a variety of factors, including how much of the substance a person is exposed to, how they are exposed, and for how long.

Why would I leave windows if Linux isn’t offering anything better?

Because Linux offers an ad-free experience, whereas Windows offers a free ads experience.

Eh, it’s the same on the Android side of the fence. There are big and small features that Google has been comically slow to crib from iOS.

I’ve definitely said “fucking finally” to things like overflow scrolling animations,

Those things like overflow scrolling, keyboard peak, etc… were only held back because Apple would patent it prevent it from being put into Android and would file frivolous lawsuits against other phone manufacturers to try and get them not to use them, even when some android variants already had it built in before apple patented it in the first place. (I still facepalm at apple trying to sue others over a rounded rectangle shaped phone)

And those patents lawsuits only stopped because other phone companies called bullshit and started threatening apple with their own patents.

and the “wild” idea that users should get 5+ of major OS releases.

TL;DR on this point: not much of an issue anymore.

This isn’t an android/iOS thing, it’s a manufacturer thing. If a chip isn’t supported by it’s manufacturer, then no software on it can be supported. Different manufacturers had different support windows, but Qualcomm became notorious for making chips, then only supporting them for 2 years so they could sell a new “supported” one (and watch the money roll in). Once they saw other the larger players getting pissed off and poking around with the idea of making their own chips, Qualcomm quickly decided that they could support their chips for longer. Now they have to since both Google and Samsung have made public promises for 5-7 year support cycles. Of course, that hasn’t stopped other phones from already reaching 7 years of official support before. (A notable example being Fairphone 2 who used a Qualcomm chip while they were still in their shitty behaviour phase and managed to support it for 7 years, 2 years Qualcomm support then 5 years of their own support despite Qualcomm.)

Also, when Google was pissed at Qualcomm they decided to start modularising their OS and pulling chunks out of it out of needing direct hardware support. This means that even if chip support were to stop, it would only affect the background / lowest-level-invisible-to-the-user parts of the OS, and all the user visible parts of the OS could be updated independently (starting with Project Treble, and going all out with Project Mainline). This basically means that entire chunks of the OS can be updated the same way an app can be, early 2010 Qualcomm companies be damned.

This also has the weird thing of android not really being a “version” per se, one phone might have different components of Android 10/11/12/13/14/etc… running at the same time. The components themselves have their own versions.

I’m not sure what the downsides are here…

Your question:

what things did the LHC discover that have real practical applications right now other than validating some hypothesis

Is really multiple questions:

1. Is doing fundamental research with no application in mind useful?

2. Has the LHC led to practical applications usable today

The answer to question 1 is yes.

There’s different types of research programs made to target different goals. Some aim for short or medium term applications, and others are just pure fundamental research.

Just because pure research doesn’t have an application in mind, doesn’t mean it’s not useful. The application isn’t the goal, the expansion of our knowledge base is. Everyone who ever thought up of an application for something did so based on their own knowledge base. If the knowledge base never expands, then we run out of applications to think up. This is why pure research is useful.

And all of history supports this:

• The discovers of rays shooting off cathode-ray-tubes in the 1800s were just doing pure research and had no idea it would lead to TVs
• particle accelerator research lead to invention of cat scans
• chemists trying to research heavier elements leading to the discovery of nuclear fission, leading to nuclear power
• electrolysis research lead to the invention of lead (and rechargeable) batteries
• etc…

The answer to question 2 is also yes:

The obvious ones are:

• improved manufacturing processes
• improved supercooled superconductors
• improved large scale vacuum chambers
• Improved data processing
• Trained a new cohort of experienced scientists/engineers/workers/etc (who can now work on new projects outside of the LHC)

The solution is simple: Remove the dagger mid-combat.

You could make the dagger too hot to hold and it falls out of reach (off a mountain, into rushing water, etc…)

You could make the dagger dissolve away (through lava, acid, being eaten, etc…)

You could make something take the dagger (disarming, stealing, etc…)

A hag/genie/etc could place a curse on the PC (holding anything makes them experience immense pain and drop what they’re holding, anything dagger they hold is turned into a spoon, etc…)

I have yet to be given an example of something a “general” intelligence would be able to do that an LLM can’t do.

Presenting…

Something a general intelligence can do that an LLM can’t do:

Play chess: https://www.youtube.com/watch?v=kvTs_nbc8Eg

Why can’t it play it? Because LLM’s don’t have memory, so they can’t work with logic. They are the same as the little “next word predictor” in your phone’s keyboard. It just says what it thinks is the most probable next word based on previous words, it’s not actually thinking or understanding anything. So instead, we get moves that don’t make sense or are completely invalid.

Did you purge and update your filters?

Note: I’m not talking about turning filters off then back on, I’m talking about updating the version of each filter itself.

“the means” in this case would be authoritarian repression.

“The means” always has to be something bad for the “ends” to try and justify reaching for “the means”.

I’m curious why people would downvote a request for port forwarding?

Also, you can buy Tic Tacs from any newsagent or gas station.

I’m confused and didn’t understand this point.

Both of the screenshots used in the article show the street names.

Every street is shown on the zoomed in screenshot, and every major street is shown on the zoomed out screenshot.

I wonder if they’re lying about this. Maybe the fans are super loud or something and they didn’t want the reporter to know.

That’s far too conspiratorial for me. Loud fans in an engineering sample aren’t a reason to break a fan.

A fast fan blade on a laptop would snap easily if it was handled, which is exactly what would be happening on both a laptop where assembling and disassembling it is a feature and a laptop being actively tested.

If it was a blade that broke, that wouldn’t stop the fan from working, so it was probably the servo, power, or bearings which is exactly what you’d expect to find broken in an engineering sample. Why? Because engineering samples almost always have issues in them. That’s the whole point of the samples, to find out what the issues are so they can be fixed before mass manufacture.

I can’t see your comment about heavy dev and testing.

I’m curious about what exactly is chewing up that much RAM. Do you have a ridiculous amount of containers running? Or a big ram disk or something?