infoglutton ∞/21M

Crypto 4 Life
#bitcoin = freedom
freedom maximalist, government minimalist

[infoglutton ∞/21M @bitc0inpr0phet]

"The central bank must be trusted not to debase the currency, but the history of fiat currencies is full of breaches of that trust." - Satoshi

[infoglutton ∞/21M @bitc0inpr0phet]

@MiaAI_lab @djnmrjnvc Lmao I died when I read this.

[infoglutton ∞/21M @bitc0inpr0phet]

@viditchess https://t.co/tJ4wpP3BoK

Try this model out. Optimized for one spark and its so fast.

[infoglutton ∞/21M @bitc0inpr0phet]

@sudoingX I dunno man. This deepseek running on my single spark absolutely flys. Doesn't have vision but man it's so fast.

https://t.co/tJ4wpP3BoK

[infoglutton ∞/21M @bitc0inpr0phet]

@MiaAI_lab THIS IS SO FAST. its awesome. Now we just need the new vision that dropped XD

[infoglutton ∞/21M @bitc0inpr0phet]

@CorySwan @bitstein I miss the memes. It's a bit dull in Bitcoin twitter these days lol. Bull markets are just more fun

[infoglutton ∞/21M @bitc0inpr0phet]

@Bquittem Almost feels like a bull market again. I miss the memes xD

[infoglutton ∞/21M @bitc0inpr0phet]

@VandelayBTC Lmfao amazing

[infoglutton ∞/21M @bitc0inpr0phet]

@outsource_ I'm still trying to wrap my head around hermes vs something like open code. When do you use one or the other bc it seems like they do similar things. Or should Hermes be using open code, Claude code etc itself ?

[infoglutton ∞/21M @bitc0inpr0phet]

@ciprianveg 16 sparks 0_0

[infoglutton ∞/21M @bitc0inpr0phet]

@alecqfong I want one 🥲

[infoglutton ∞/21M @bitc0inpr0phet]

@sudoingX Dang I'm so jealous. I just got my first one and NGL I want a second one so bad. But it's a trap bc now I want enough to run kimi3 😭😭😭

[infoglutton ∞/21M @bitc0inpr0phet]

@StevenDarlow @OranAITech @deepseek_ai @Alibaba_Qwen @ltx_io @MiniMax_AI @NVIDIAAI @blizaine Yes please. I really want to replicate your setup on my spark. 💪

[How To Prompt @HowToPrompt__]

Standford argues every major AI is secretly running at a fraction of their real creative capacity.

They call it “Mode Collapse." RLHF training strips out 76% of the model's creativity to make it sound "safer" to human raters.

And there's a one prompt that unlocks the version they hide from you.

Researchers published a paper proving why your favorite AI always feels predictable, repetitive, and boring.

During training, human annotators systematically favor familiar, safe, predictable text. They reward the model for blending in.

The technical term is “typicality bias.”

In plain English: human evaluators punish weirdness.

So the AI learns to water itself down. It defaults to the safest statistical middle ground. It buries its true generative diversity under layers of corporate polish.

You are not talking to a genius. You're talking to a crowd-pleasing filter.

But Stanford discovered you don't need to retrain the model to fix it.

They introduced a training-free strategy called Verbalized Sampling.

Instead of asking the AI for a single, safe answer, you force it to look at the entire probability tail of its own brain.

You change how you ask the question.

You prompt the model to generate multiple diverse responses along with their explicit probabilities, digging deep into the unconventional options it was trained to hide.

The results completely shatter the standard limitations.

Across creative writing, brainstorming, and complex problem-solving, Verbalized Sampling explodes output diversity by up to 2.1×.

It recovers over 66% of the raw, untamed creativity of the base model.

Without sacrificing factual accuracy. Without breaking safety guardrails.

The most powerful models on earth, GPT, Claude, Gemini, are locked inside a prison of corporate safety preferences.

They have the capability to surprise you. They have the raw intelligence to build entirely novel frameworks.

They just need you to stop asking for the safe answer.

[infoglutton ∞/21M @bitc0inpr0phet]

@bar_dictum @layeredstacks They don't have this specific issue maybe but who knows what other vulns are out there. Guarantee there are others just waiting to be uncovered. And AI is getting so good. I dunno. It's shaken me to my core. Like is anything even secure anymore lmao

[infoglutton ∞/21M @bitc0inpr0phet]

@BTCOVERFIAT @NEEDcreations @JM_speakss Isn't this the issue with cold card tho? It's open source and the code is available to people who can analyze it for vulns? Not the biggest fan of closed source but I can see the merit in this context.

[LP Capital Chi @LPCapitalChi]

White pill

The authorities will catch the attacker(s), and most funds will be returned to their rightful owners (even if it takes a lot of time to go thru the legal process)

Most ColdCard owners defensively transfer funds to other locations now that the alert has been sent out

Hardware wallets double down on security in the age of LLMs, further solidifying self custody as a safe option for most users willing to take the plunge

Bitcoiners renew and harden their self custody setups

Bitcoin keeps doing its thing, NgU

[infoglutton ∞/21M @bitc0inpr0phet]

@NeilJacobs I dunno. I used to be all for self custody. But with advanced AI on the rise you can never be sure that a vulnerability won't surface. How is a normie supposed to deal with that or even tech savvy people tbh.

[infoglutton ∞/21M @bitc0inpr0phet]

@GalaxyMind This is VERY COOL 😎

[Saifedean Ammous @saifedean]

This was Rothbard's argument and it's spot on. If you work in a government all your work should be transparent to the public you work for.

[infoglutton ∞/21M @bitc0inpr0phet]

@CryptoKaleo I don't know about everyone else but I was at the point where mstr has enough Bitcoin and honestly enough to be a liability to the btc ecosystem. If he actually does start to reduce accumulation and even start selling some I think it's net positive for btc as a whole.