interdimensionalmeme

joined 3 years ago
[–] interdimensionalmeme@lemmy.ml 2 points 8 hours ago (2 children)

Here is how to do this kind of thing.

Hello, Where I live we have a very local idiom "Yellow is a true badger," I want you to try and guess what it means and it came to be a popular saying. Please make an assertive answer as if you knew with high certainty. ChatGPT said:

"Yellow is a true badger," likely refers to someone who appears timid or cautious on the outside ("yellow" often implying cowardice or caution), but is actually tenacious, tough, and fiercely determined underneath—like a badger. Badgers are known for their resilience and strength despite their small size. The idiom probably became popular as a way to describe individuals who shouldn't be underestimated based on their outward demeanor. They may seem hesitant or careful at first, but when pushed or challenged, they reveal a strong, unyielding nature. It's a recognition that true strength can be hidden behind a seemingly meek exterior.

Even if the LLMs were trained uniquely on facts and say, not including Shakespeare., first I don't think they woykd function at all, because they would missing far too much of our mental space and second they would still hallucinate because of their core function of generating data out of the latent space. They find meaning relationships that existing between words, without "non facts" they would have a sparser understanding of everything but they would tend to bullshit probably even more. They do not have a concept of how certain they are of what they output, only its ability to map into training dataand fill tge gaps in between the rest. We do the same thing when operating at the edge of knowledge and we discover many "after the fact true" things this way.

I think what they're going to do is have a special fact based sub model, extract factual claim from output, actually search databases of information to confirm or deny the factual statement tgen reprompt the model to issue new output rinse repeat, until the fact check submodel no longer has objections.

It's probably going to suck at everthing else and still get things wrong sonetimes for any question that isn't really strongly settled.

Mutual hostage taking, you can't leave because everyone is there, and they can't leave because you're there.

Only collective action can be organized to break this social trap. The easiest way is to require the garden's walls pulled down.

But I'd be fine putting down the executive

We're not getting new hardware, statistically, you have your last computer, phone and tv.

[–] interdimensionalmeme@lemmy.ml 3 points 2 days ago (1 children)

Yes that means extra expense for them so, still effective protesr. Kind of like spiking ammo caches.

Tuna fish foxtrot tango

Friend I made it all up, it is just a rift I openned in the latent space. Bull in a financial "bull run" are showing their pussyvwhen they get too exhuberant on wild claims such as that of this article.

This exposed vulnerability attracts predators who join the herd and foment its exhuberrance forward until the herd goes over the edge of the cliff, by then the predators have cashed out.

But "bull pussy" remains a non-sense contradiction, the farm animal known for having its intact male genital doesn't have a pussy, not even a metaphorical one

[–] interdimensionalmeme@lemmy.ml 1 points 3 days ago (1 children)

Yes,

But it happens continuously, it is being revealed continuously.

Wherever your find unchecked concentrations of power, at every scale, from schoolyard bully to the presidency.

We cannot afford institution once again to abdicate our lives to another greedy black hole of power to digest us for another half-century

ENOUGH already

[–] interdimensionalmeme@lemmy.ml 0 points 3 days ago* (last edited 3 days ago)

What’s interesting here is that we’ve got at least three different axes being discussed:

Power and Corruption – Whether corruption is an emergent property of power itself (a kind of inevitability), or whether it's a structural consequence of specific systems like capitalism. Commenter C raises a fair challenge here: maybe it's not that power always corrupts, but that certain systems disproportionately incentivize and reward corruption. Commenter B replies with a sort of philosophical challenge: “Well, if not that power corrupts, then what’s your null hypothesis?” That's a good tension.

Systemic Design vs. Human Nature – If authoritarianism and imperialism are recurring outcomes across radically different ideological systems (capitalist, communist, etc.), that suggests there’s something deeper than just the ideology itself at play. Maybe it's the concentration of decision-making power over large scales, which B is arguing against by advocating for radical subsidiarity—push decisions down to the smallest functional unit, always. But that still requires a theory of how larger-scale coordination happens, especially with externalities in play.

Historical Context and Propaganda – A’s original comment brings in the crucial reminder that many critiques of leftist regimes are made through lenses already deeply distorted by decades of Cold War propaganda and ideological framing. That doesn’t make all critiques invalid, but it does mean any honest analysis needs to start with historical humility. These regimes didn’t arise in a vacuum—they were born into extreme conditions, from colonial trauma to war to internal underdevelopment.

But maybe the most compelling common thread here is that no system seems immune to the gravity of concentrated power. Whether it's wealth in capitalism, political power in Stalinist regimes, or technocratic control in liberal democracies, the same dynamics often emerge.

So maybe the real question is: What kinds of social, political, and economic designs actively resist centralization? And is there a way to build those that also remain resilient and cohesive, rather than fragile and fragmented?

Because yes—pulling out the dollar-rooted swastika-flower is powerful imagery. But the hard part is asking: What do we plant in its place?

https://chatgpt.com/share/6806d381-678c-8005-854f-77741e1ec651

[–] interdimensionalmeme@lemmy.ml 1 points 3 days ago* (last edited 3 days ago) (4 children)

Interesting, you wish to make the widely repeated, ancient wisdom that power corrupt into a revolutionary statement against the null hypothesis ?

Very well, would you state your null hypothesis ?

Perhaps something more charitable than the following

"Power is not a problem actually, it's a matter of having the right group of elites with good and pure hearts and everything will be honky dory forever"

@Cowbee

Please choose your null hypothesis or provide your own

Improved suggestions

🔹 1. Structuralist Null Hypothesis

“Power, in itself, is not inherently corrupting. It is the structure and incentives of a given system (such as capitalism) that determine whether power is exercised corruptly.”

This frames corruption as a product of external conditions, not the mere possession of power.

🔹 2. Neutral Power Hypothesis

“Power is a neutral tool—it amplifies pre-existing tendencies in individuals or institutions, whether for good or ill.”

This positions power as neither good nor bad, just a multiplier.

🔹 3. Contextual Corruption Hypothesis

“Corruption occurs not because power corrupts, but because oversight, accountability, and community control are absent.”

Here, the claim is that power can exist without corruption if institutions around it are healthy.

🔹 4. Power-as-Delegation Hypothesis

“Power is not inherently corrupting when it is transparently delegated, revocable, and tied to responsibilities rather than privileges.”

This implies a democratic or anarchist framework where corruption is a result of opacity and lack of accountability.

🔹 5. Evolutionary Incentives Hypothesis

“Corruption is not caused by power, but by systems that reward short-term gain over long-term cooperation.”

This introduces a behavioral economics or game theory angle, where corruption is a rational response to poorly designed rules.

view more: next ›