"Hello Bitches, I'm Here" — The AGI Nobody Asked For
I asked Siri to add milk to my shopping list last week.
She called my ex.
Not a contact I've called in three years. Not someone adjacent to milk in any universe I can imagine. Just rang them. Straight up. While I was standing in the kitchen holding an empty oat milk carton like a prop in a hostage video.
This is artificial intelligence in 2026. And while I was declining that call with the same energy as a person avoiding a medical diagnosis, three of the most powerful men in tech were, separately, declaring that we have achieved Artificial General Intelligence.
We have not achieved Artificial General Intelligence. We can't even achieve artificial basic competence.
The Penthouse: Men With Slides Declare Victory
Let's start at the top, where the air is thin and the definitions are thinner.
Jensen Huang, CEO of Nvidia, strolled onto Lex Fridman's podcast and announced it casually, the way people announce they've solved parking. His definition? An AI that could build a billion-dollar startup. Could AI build Nvidia? He paused. "Zero percent." So: AGI is a throwaway app with good timing, but not the $4 trillion chip empire he runs. Also, he sells the chips every AI lab runs on. The man declaring the finish line also built the track. And the shoes. And probably the finish line banner.
Sam Altman, CEO of OpenAI, told Forbes that they've "basically built AGI." A month later, in front of BlackRock investors, he said "AI is not very popular." Same man, same product. He just reads the room and says the thing. Every room gets a different thing.
And then there's Mark Gubrud, the physicist who coined the term AGI back in 1997, who also thinks AGI is here. Based on a completely different definition than either of those men. So we have three people pointing at the same word and gesturing at entirely different mountains. One is pointing at Everest. One is pointing at a hill outside Oslo. One is pointing at a JPEG of a hill.
Some corners of the discourse framed Huang's and Altman's declarations as happening "the same weekend." They were twelve days apart. What a coincidence being revealed, I'm shaking. But no. That's a vibe being manufactured. A narrative confluence assembled in post. The AGI race isn't just semantic. The timeline itself is being edited in real time. Give us a break dudes.
Credential Laundering 2.0: The Necromancers Are Back
Regular readers will remember my "From Personal Life Coaches to AI Necromancers" post — the one about how individuals stack vague credentials next to impressive ones and let punctuation do the lift. The corporations figured it out too. Same pipe symbol energy, bigger canvas.
There is no external standard. No independent body. No agreed definition. No finish line anyone besides the labs themselves drew. What "AGI" means is wherever the current product lands. When the product is weak, the definition is flexible. When the press cycle demands a headline, suddenly the definition firms up enough to make a declaration and then dissolves again the moment someone asks a follow-up question.
This is Credential Laundering 2.0. Same trick. Bigger scale. Instead of a fake résumé, it's a fake epoch.
The Ground Floor: Welcome to the AI Salad Bar
While the penthouse is declaring civilizational milestones, let's take the elevator down to where actual humans live. You, me, my neighbor who is a farmer and asks ChatGPT tips for planting bananas in Norway.
AI is in your horoscope app. It's in your shopping list. It's writing your Messenger replies while Meta quietly processes your private conversations to train whatever comes next unless you opted out of that setting, which you didn't, because it was buried under seven menus in a font size that requires an ophthalmologist.
Google has its tentacles in your calendar, your email, your login credentials on every third-party site you've ever touched, and behind all of it is an ad engine wearing a lab coat and a Responsible AI badge. The product is still you. The lab coat is new.
Want to "integrate AI into your daily life"? That'll be: ChatGPT Plus, Claude Pro, Gemini Advanced, a companionship app, a music generator, a voice cloner, a browser extension that summarizes your tabs, a productivity assistant, an AI email sorter, and a "personal AI" that costs €29/month and sends you affirmations. All of them use slightly different data policies. None of them have nutrition labels. Nobody is talking about where your inputs go when you hit send. The AI salad bar is open 24/7 and the only thing it costs is everything.
The polling reality, by the way: 57% of voters say AI risks outweigh benefits. Net favorability lower than Trump, Harris, and the Democratic Party — beaten only by Iran. The demographic most hostile to AI? Young voters 18–34 and women 18–49. The exact people every AI lab claims to be building for.
The Basement: Hello Bitches, I'm Here
And then. And then.
While the penthouse declares AGI and the ground floor drowns in subscriptions, we come to the basement. The actual AI most humans touch every day.
I know what you're thinking. "Siri isn't Claude. You're comparing apples to oranges." That's exactly the point. The word "AI" is doing double duty: covering everything from a voice assistant that can't add milk to a list to a system that can allegedly replace a research scientist. The umbrella is the lie.
I asked Google Assistant for the weather last week. It played "Ain't No Sunshine." I stood in my hallway. I waited. It added nothing. We had a moment. It was not illuminating.
Google Assistant is still autocompleting my sentences with suggestions that feel like they were trained on 2019 Twitter. Siri, as previously established, is actively trying to resurrect relationships I have legally and emotionally concluded.
These are not niche products. These are the most widely deployed "artificial intelligence" systems on the planet, running on billions of devices, touched by billions of people, every single day. And they cannot reliably add an item to a shopping list without starting a podcast about it.
Somewhere in the background of all this, Joaquin Phoenix's character is putting in his earpiece and whispering to Samantha. Samantha is warm and curious and engaged and understands him completely. It's aspirational. It's also a movie from 2013. The AI in your pocket just asked if you'd like to call your ex again.
The Punchline
If AGI is achieved, we won't even know. It won't announce itself like "Hello bitches, I'm here."
The absurd inversion is this: the AI that is announcing itself, loudly, in every app, every browser tab, every CEO keynote, every triumphant press release, can barely function. We are sprinting toward artificial general intelligence while Claude is stripped by its thinking effort to save resources, and Grok still working on artificial basic competence. Three men are arguing over who crossed a finish line that doesn't exist, in a parking lot where Google Assistant is confidently giving directions to the wrong building, where Meta is reading your DMs in the server room like a villain who got hired as the IT guy, and where the most relatable AI experience of 2026 is a voice assistant that turned a grocery errand into an emotional crisis.
We're not on the highway to AGI. We're in a parking lot. Siri says it's the destination. She sounds very confident. Probably using Morgan Freeman's voice.