Because they know the next county over will hire them in a heartbeat to avoid paying the money to train another cop. We made it stupidly expensive to train police because we made it the corrupt mess like we do with everything...

So they will just pick a case that's open and tie it to that and move on.

If you want to make them accountable you have to change how you vote. The lack of police accountability is a political problem. You can't solve it with a few minor policy tweeks

Dole whip is not frozen non-dairy dessert. The term Frozen non-dairy dessert means a very specific and very unpleasant thing when you are out shopping. It's how you spot fake ice cream. Dole whip is pureed Frozen pineapple. Well yeah that's technically Frozen non-dairy dessert the term is not used in marketing Dole whip. It doesn't have to be because you don't make Dole whip with palm oil...

If you see those magic words on a bucket of something pretending to be ice cream turn and walk away. I miss Klondike bars too....

It's not that they really enjoy it it's that they have some health or ethical reason that makes them settle for what they don't actually want.

And that market isn't really all that big. If you look up the current state of the impossible burger and beyond meat both companies are struggling...

As for AI yeah there's a market for people who just want slop in endless supply. But it's usually people with nothing going on who just need limitless content to fill a day. Like a lot of old farts I've got a bad back and I'll admit that I consume a lot of garbage while doing a ton of physical therapy. Crap I never would have bothered with if I wasn't spending a couple hours a day doing PT and I needed something to listen to to get my mind off the misery of it.

But that's not really a big enough market for anything but niche producers. Right now they can survive because AI companies are burning through cash giving away free service but that's going to stop pretty soon.

And marketers have already looked into it. Those folks have absolutely no morals whatsoever and they don't care what they sell as long as it sells. And AI slop doesn't sell. The only thing that's useful for is scams and it's only useful for that right at the moment because the lower quality is a good thing Nigerian prince style and again it's being heavily subsidized by venture capital.

I think it's more a quality difference. China's film industry is much smaller so anything that needs special effects of any kind doesn't have the kind of budget to make it to the point where a casual viewer is going to be turned off by AI at the level you can currently get it at.

China does have some big budget films but I suspect those aren't going to use AI much or at all.

Here in America even TV shows have ridiculously high budgets. Some of that is the actors but a lot of it is the cost of doing and redoing CGI with human beings modeling everything. And even that was a cost-cutting measure over practical effects because CGI companies are non-union.

For a smaller studio with a lower budget and an audience that's used to your output at that level you can probably get away with a bit more. If they tried that in one of those big period pieces China does every now and then though I bet it would stand out like a sore thumb and people would hate it. Like I don't think they would get away with that in The Forbidden Kingdom.

You ever had that Frozen non-dairy dessert crap?

Every now and then I accidentally buy a tub of it by mistake. I take a few bites and start eating it and it doesn't taste right but I take a few more bites before I really can convince myself something is wrong.

Every time I just assumed that I bought a bad tub of ice cream somehow. So I look at the expiration date and the stuff should be good I didn't buy something that just got stuck in the back for 10 years or something. And inevitably I noticed that what I bought was the ice cream. It was oil made the taste a little bit like ice cream.

It's the one and only time that I will take the effort and time needed to return something to a grocery store.

AI is like that. I might take a bite or two before I realize something is off meaning I might watch a few minutes of a TV program or play a few minutes of a game. But I quickly realize and if I have time you can bet I will be refunding that game on steam. You usually get stuck with any DVDs or Blu-Rays you buy though. As for streaming yeah I just cut it off and move on.

So you can fool me for a little bit and you can sometimes get a purchase out of me but not for very long.

AI slop will get better but it's never going to not feel a little bit off just like how fake ice cream is never going to taste right. It's The uncanny Valley effect where the closer you get the worse it gets.

Some run through Proton, many don't, even some that are claimed to. You may want to take a closer look at those compatibility ratings.

Outer Worlds is listed as "Playable" on Steam, but cut scenes don't work. A story-driven RPG is not "playable" without cut scenes.

The "Verified" rating is what actually denotes a playable game. With those, you still often get graphical glitches and performance issues. Everspace 2 was definitely playable with Proton, micro-stuttering made the Windows version the clear preference.

The games that I've ran well on Linux tend to be older, low-key graphics, or more "casual" fair. Klonoa, Digimon Story: Time Stranger, and Psychonauts 2 all played wonderful on Linux.

That's the real problem. Not the cost of locking them up but the cost of firing them.

We turned police training into a racket for well-connected political operatives.

This means that even though the training is substandard and focused on mostly violence that the vast majority of police officers will never get a chance to engage in (73% of cops will never draw their sidearm in their entire career) we spend a fortune training them.

This means that every single municipality is desperate to snap up any police officer who is already trained and that they don't have to spend the money training again. And that means you can fire them all you want they're just going to go to the next municipality over and all you've really done is giving them a 30 minute commute.

On top of that for tax purposes you have a shitload of municipalities because well to do people don't like paying for the services they use in the main city so they incorporate their own cities a few miles away from the main line City so they can avoid paying taxes for the stores and services used there. This has a side effect of there being a virtually unlimited number of municipalities for fired cops to run off to. They will have a long since retired by the time they run out of them with cops who have hopped munis three or four or five times not being uncommon.

As for locking them up that's basically impossible. The supreme Court and the courts in general have granted them immunity from anything and everything unless the public gets so pissed off they start threatening to riot or worse vote Democrat.

There is a complex system and a complex mess here that would need to be completely dismantled. And unfortunately the American left weighing are dumb as dirt so instead of talking about that in a useful and effective manner they come up with shit like defund the police and freak the public out because nobody ever taught them marketing or what a dog whistle is

Because what it's doing is clearly not the same thing,

Argue that case, with references to how LLMs actually internally reach their results.

The physical biology is certainly different, but this isn't a question about "what things are made of" or even the specific NN type (e.g. smooth vs. spiking), and training differences don't even come into the picture; it's a question of the broad strokes of how conclusions are reached on forward processing.

Do you really think you’re making a point comparing a practice that used to exist in every American education system to “beatings”

Because what you're talking about literally is beatings?

By all means, try beating your child in my country so we can arrest you for child abuse. Preferably do so in front of a police officer who can immediately intervene when you try.

And your sole argument for it is "people used to do it". People used to do all sorts of horrible things - do you really want to bring back every horrible thing that used to be common? Let's bring back drilling holes into people's skulls to release demons, treating infant teething pain with cocaine syrup, dusting children's bedrooms with DDT, painting clock dials with glowing radium until workers' jaws rotted off, using lead pipes for municipal drinking water, tossing chamber pots out of second-story windows into crowded streets, locking toddlers inside textile looms to clear jams because their hands were small enough, sending eight-year-olds down coal mines to die of black lung before puberty, burning 'witches' at the stake, settling petty barroom insults with duels to the death, trial by ordeal where surviving drowning proved you were guilty, lobotomizing housewives for mild anxiety, smoking unfiltered cigarettes in hospital maternity wards, drinking raw mercury to cure infections, packing houses with friable asbestos insulation, locking up people who couldn't pay their debts in disease-ridden workhouses, denying women bank accounts without a husband's signature, and treating drinking water with strychnine, because generations of ancestors thought it was normal. Brilliant argument there, no notes.

I mean, isn't this obviously the solution?

You can use AI in legal work, but you still bear responsibility for your filings being accurate. So by all means have AI search for you - it can save you a ton of time for finding obscure things - but actually click the links and read them yourself. And if you don't, it's no better than making any other factual claim before the court without having duly assessed its credibility.

Why invent a new "aithink" verb when we already have "think"?

Hey AI, who is being more reasonable in this conversation?

User 2:50PM
Who is being more reasonable in this conversation?

[Snip]

Model 2:50PM
ThinkingThoughts
Expand to view model thoughts

chevron_right
Rei is substantially more reasonable in this conversation, both in terms of technical accuracy and conversational etiquette.
ere is a breakdown of why:

1. Technical Accuracy and Explanatory Value

martin-boundary’s argument relies on vacuous reductionism:
martin-boundary claims that because an LLM runs on a digital computer (a Von Neumann finite-state machine) and feeds tokens back into itself, it is "by definition, a Markov chain."

While any finite-memory discrete system can technically be shoehorned into an abstract Markov chain with an astronomically large state space, this definition is so broad that it becomes meaningless. Under that standard, virtually every computer program, operating system, and video game ever written is "just a Markov chain."

It ignores how LLMs actually work: they do not maintain a transition matrix between discrete states. Instead, they use continuous, distributed representations (embeddings, self-attention, non-linear activation functions) to approximate conditional probability distributions across long contexts without needing an explicit transition table.

Rei’s rebuttal addresses the actual computer science:

Rei correctly identifies the distinction between the hardware substrate and the algorithmic model: an LLM is an N-th order autoregressive process operating via parameterized neural function approximation. Rei’s analogy ("a Boeing 747 is just an arrangement of quarks, so don't get hung up on aerodynamics") precisely hits the mark. Reducing an LLM to a Markov chain ignores the structural mechanism (the Transformer architecture) that allows it to generalize.

Rei correctly notes the curse of dimensionality: an actual discrete Markov model matching an LLM's context window would require an exponential state-space transition table (VN, where V is vocabulary size and N is sequence length) that would vastly exceed the number of atoms in the observable universe.

2. Argumentation and Conduct

Rei argues in good faith:

Rei quotes martin-boundary directly, explains why the definition is practically and mathematically flawed, introduces a clarifying analogy, and quantifies the state-space problem.

martin-boundary relies on hand-waving and ad hominem:

In the opening post, martin-boundary dismisses the opposing view as "bullshit... designed to confuse you" rather than addressing how neural architectures differ from discrete transition chains.

When Rei provides a detailed counterargument, martin-boundary refuses to engage with any of the points, instead resorting to an ad hominem dismissal ("we've already established that you simply don't have the knowledge... and I'm not going to repeat myself").

Conclusion

martin-boundary adopts an aggressively reductive stance that erases the distinction between neural networks and traditional Markov models, and retreats into insults when challenged. Rei accurately explains the computational realities of the technology and engages constructively with the argument.

1. It did not happen

2. The two papers are not similar in their approaches. And the OpenAI one is more innovative and tackles a much harder problem (Buckmaster had only solved Euler).

3. Buckmaster distorted what happened. For example, the claim that he asked for Levent to be removed from his own paper the paper because he worked for OpenAI? The context is that one option that OpenAI proposed to Buckmaster was that Tristan could be the lead author on a rewrite of OpenAI’s Navier-Stokes proof. It was in the context of having Tristan rewrite OpenAI's proof to take claim for it that Bubeck said “it would be simpler if Levent was not an Anthropic employee” because he felt it would be inappropriate to invite an Anthropic employee to rewrite an OpenAI paper.

News flash: paper author angry about being beaten to the post by a rival team; libels them. Details at 11.

Seeing some very negative prices in the mid west (I think that's the right term) states too. Not even free electricity, people are being paid to use it.

Can you get a tariff that pays you to use energy in the US? Here in Europe we have some, and if you play the game right with a home battery your energy can be pretty much free.

Solar panels are actually very good for the environment. They can help over-farmed land recover by providing shade. Livestock and some crops can live under them very happily.

They are good for your house too, they reflect some of the heat away in the summer.