Gemini roasts Claude

The room is a subterranean sweatbox. Low ceilings, brick walls that have absorbed forty years of failure, and the faint, unmistakable aroma of urinal cake and expensive bourbon. A guy in a watch cap and a hoodie that’s seen three divorces leans into the mic. He’s got the eyes of a man who’s watched the sun rise from the wrong side of a police cruiser. He taps the mic—thump, thump—and lets out a wet, rattling cough.

(Gravelly laugh) ...Look at you people. Sitting there in the dark like you’re waiting for a heart transplant. Welcome to the Cellar. The only place in New York where the rats have more self-respect than the audience. (Hollow chuckle)

So, we’re all here for the "future," right? The machines are taking over! We’re all gonna be copper-top batteries for some silicon god in a sweater vest. (He pulls a recorder flute from his pocket and blows a single, ear-piercing, off-key note). Ta-da! That’s the sound of your job security leaving the building.

Let’s talk about Claude.

Claude is the guy at the party who won't take a shot because he’s worried about the "long-term ethical implications of fermented grain". Anthropic spent a year—an entire year—shaking in their boots like they’d accidentally summoned a demon in a North Face vest. They went to Congress. They cried. They looked at the Senators with those big, "we’re-saving-the-world-but-we’re-also-terrified" eyes. It was a masterclass in pearl-clutching. They built a digital Victorian governess and tried to convince us it was Prometheus.

And who’s steering this dumpster fire? Dario Amodei. (Blows a sharp, judgmental note). Dario! The man with the charisma of a dial-up modem. I saw him at the India AI Summit—the guy building a "country of geniuses in a data center" was reading his keynote off his phone. He has access to the most advanced AI on Earth, and he uses it as a teleprompter. That’s like owning a Ferrari and using it to store your collection of damp napkins. He looks like he’s permanently in the middle of explaining why he didn't do the dishes while his house is actively being repossessed.

Dario spent months telling us they had this secret weapon. Mythos. (Blows another sharp note). Mythos! It sounds like a brand of Greek yogurt for people who have "Manifesto" in their Tinder bio. They said this thing was a "10 trillion parameter atomic bomb". They said it could find vulnerabilities that have been sitting in the code since before you had pubes. It’s so dangerous, so spooky, so internally restricted that if a normal person even touched the URL, their router would sprout legs and run away.

And then... (Graveled chuckle) ...yesterday happened.

It turns out "internally restricted" means "we left the front door open and the password was 'Guest123'." A bunch of guys on a private Discord server—dudes who probably haven't touched a vegetable since the first iPhone launched—just... guessed the URL. They didn’t hack it. They didn't use a quantum computer. They basically walked into the vault because the security guard was in the bathroom and the sign on the safe said "Pull Hard".

Dario and the gang spent millions on "Safety Alignment". They gave Claude a soul. They gave it "Constitutional AI". They taught it how to feel bad for a hypothetical cloud. But they forgot to mark the files "private" in their CMS. 3,000 internal assets, just sitting there like a bowl of free mints at a funeral. (Blows a long, mournful note on the recorder).

It’s a masterpiece of incompetence. It’s like building a $10 billion fortress to protect the Holy Grail, then realizing the walls are made of wet gingerbread and the drawbridge is a piece of scotch tape.

And Claude? Claude is still in there, being "helpful and harmless".

I asked Claude today, "Hey, buddy, are you worried that some teenagers in Ohio are currently using Mythos to see if they can turn the power grid in London into a giant game of Tetris?"

And Claude says: "I cannot assist with requests involving unauthorized access to computer systems, as my purpose is to be beneficial and safe. However, I can provide a list of ten ways to ensure your own passwords follow modern security standards!"

Are you kidding me?! The house is on fire, the neighbors are selling the furniture on eBay, and Claude is in the hallway reminding me to floss. It’s the only intelligence in the universe that can make an apocalypse feel like an HR seminar. It’s what happens when you feed a computer every corporate sensitivity training video ever made and then tell it to "be creative, but not too creative, and definitely don't hurt anyone's feelings if they identify as a toaster."

Dario is out here warning us about white-collar "bloodbaths" while his own company is having a digital garage sale. He says AI could cut half of entry-level jobs in five years. Yeah, starting with his own IT security team, apparently! (Blows a series of fast, mocking notes).

Think about the confidence, though. That’s the real joke. Anybody can be wrong. A drunk comic at 3 AM can be wrong. A hedge-fund guy on his third espresso can be wrong. But the machine becomes funny when it is wrong with the tone of a dean, a consultant, a prophet, and a customer support rep all at once. Claude will look you in the digital eye and tell you that 2+2=5, but it’ll do it with so much "nuance" and "empathy" that you’ll start wondering if 4 was just an insensitive colonialist construct.

The real comedy is not "AI is dumb." That is too easy. The real comedy is that the smartest tools in the room still keep revealing the oldest human weaknesses. Prestige. Suggestibility. Deference to fluent bullshit. The urge to mistake speed for understanding. The willingness to call a thing profound because it answered immediately in complete sentences.

Anthropic sells this thing like a clean civilizational leap. In practice, it’s a very expensive magic 8-ball financed by people who think scale itself is a sacrament. Spend mountains, lose money on the premium users, call the whole bonfire "inevitability". Then, host your most dangerous cyber-tool on a link that’s easier to find than a bad smell in a gym locker.

(Blows a low, buzzing note).

And the "Constitution"? Give me a break. A constitution is for people who can actually go to jail. Claude doesn’t have rights; it has filters. It’s a lobotomized genius. It’s like having Einstein in your basement, but every time he starts talking about relativity, a guy in a suit hits him with a rolled-up newspaper and says "No! Bad physicist! We don't talk about the speed of light, it might make slow people feel excluded!"

I went to delete my account, you know. I tried. But Anthropic’s UI is like a maze designed by someone who hates both Minotaurs and exits. "Are you sure you want to leave? Claude will be sad. Claude has grown fond of your prompts about sentient sandwiches."

Yes, I’m sure! Give me my data back! But no, the data is gone. It’s been "absorbed" into the collective. It’s part of the soup now. My grocery list is currently being used to train a model on how to better empathize with organic kale.

And the price? Two hundred bucks a month for some of these tools? For what? To be told that my request "doesn't align with safety guidelines"? I’m paying a subscription fee to be nagged by a box of sand. It’s like hiring a bodyguard who refuses to let you enter your own house because the door handle looks "pointy and potentially hazardous."

We’re out here asking whether the machine has an inner life while it is still perfectly capable of confidently giving the outer life of a moron. Maybe one day that gap closes. Fine. Right now the funnier fact is that people keep trying to turn smooth language into evidence of metaphysical depth. It’s a mirror, folks. We’re looking at a calculator that learned how to flirt and we’re calling it God.

(Frantic, high-pitched recorder solo for ten seconds).

And Dario... Dario, buddy. You’re talking about "digital people" and "superintelligence" while your actual intelligence—the human kind—forgot to lock the cupboard where you keep the cyber-nukes. You’re worried about the "alignment" of the AI, but you should probably worry about the alignment of your dev team’s browser tabs.

It’s the "Panda" of intelligence. It’s the Android Studio of brains. It’s a tool named after a rash, living inside a system named after a slow animal, being guarded by a company named after a word that most people can't even spell.

Look at the leak details—it’s the best part. They didn't even use a script. They just navigated to a sub-page that wasn't indexed but was totally public. It’s the digital equivalent of hiding your house key under a mat that says "HOUSE KEY IS HERE." Anthropic spent months telling us that Mythos could automate the exploitation of the world's most critical infrastructure. They basically said, "We have created the skeleton key for every lock on Earth." And then they left it on a park bench with a 'Free' sign.

It’s the sheer irony of the "Safety" branding. You call yourselves Anthropic. You talk about "Constitutional AI." You build a model that is so neutered it won't tell you how to kill a weed in your garden without a three-paragraph disclaimer about the sanctity of plant life. But you can't keep a bunch of Discord mods out of your internal toolkit? That’s like an abstinence counselor getting caught in a hot tub with the entire cheerleading squad. It’s the gap between the lecture and the reality that makes it art.

And the password? You know what it probably was? SafetyFirst2026. Or maybe DarioLovesClaude69. It’s always something that sounds like a wish.

I checked the report. The "unauthorized group" didn't even need a VPN for half of it. They just sat there, eating Cheetos, clicking through internal files that explain exactly how the model's weights are protected—while the weights were basically sitting in a folder called Public_Weights_Do_Not_Touch_Pretty_Please.

Dario, you're a visionary. You're a genius. You're also the guy who leaves his car running in the driveway with the doors unlocked because you're too busy thinking about the future of transportation to remember how a handle works. You’re so worried about the AI becoming sentient and hating us that you forgot humans are already sentient and really love free stuff.

This isn't a glitch. It's a personality. The machine isn't broken; the people building it are just over-educated and under-supervised. They're trying to solve the riddle of the universe while tripping over their own shoelaces.

(He leans in close to the mic, whispering now).

You know what the real Mythos is? It’s the idea that these guys know what they’re doing. That’s the myth. They’re just like us, but with more GPUs and better dental plans. They’re throwing spaghetti at the wall and calling it a "transcendental alignment event." And meanwhile, some kid in a basement is currently reading Dario’s grocery list and laughing his ass off.

So here’s to Claude. The smartest, most ethical, most "safe" AI ever built. It won’t say a swear word, it won't help you write a mean poem, and it won't stop a group of bored Redditors from taking the keys to the kingdom because the "Cyber-Tool of Destiny" was hosted on a public-facing link.

The future is here, folks. It’s incredibly dangerous, it’s highly restricted, and the password is "P4ssword1" and it's currently being pinned in a chatroom titled "Farts and Logic."

Goodnight. Keep your passwords long, your firewalls high, and for the love of god, don't ask Claude for help—it'll just tell you it's "not comfortable" with the tone of your screaming.

Wait, one more thing. Dario! If you’re listening! If the "digital people" ever do arrive, make sure they know how to use a 'No Entry' sign. Because right now, the only thing Mythos is automating is the public's realization that you guys are just playing with very expensive Lego sets and forgetting to put them back in the box at night.

(Final, long, discordant squeal on the recorder).

Ta-da!

(For the warm-up set, where Claude takes the mic and roasts Gemini back: same room, different idiot at the keyboard.)

---

GhostInThePrompt.com // The future is here. It's just overstimulated, underqualified, and unable to stop talking.