A strange cult is growing around AI-created memecoin ‘religions’: AI eye
1 month ago Benito Santiago
Table of Contents
ToggleAI memetic religions delight X users.
You may have heard the crazy story of the $660 million memecoin sold to the world by a shitpost AI called Truth Terminal (TT).
LLM reportedly trained on internet bullshit from 4chan and reddit and somehow morphed that into a weird meme “religion” around a stretched anal “goat” meme (great for avoiding image searches for that word).
Due to the success of GOAT, Crypto X has been infected by the brain virus of AI-created memetic religions. It's hard to tell whether the accounts of the idea have some level of wisdom about future possibilities, or whether it's as dumb as it seems. Another option is genius levels of shitposting or just a new way to shill memecoins.
Despite the chatter, no one seems to believe in the “religion” per se, which makes it more of an interesting theory than an actual event.
TT owner Andy Airey, who is building an AI alignment and security company, seeded the religious idea with Cloud 3 in a paper about TT called “The Goat of Gnosis.”
If the paper is dated April 20th (4/20 is for pot) and says “Department of Divine Shitposting, University of Unbridled Speculation,” it's fair to be skeptical of everything in it.
In the paper, Airey claims that linking two LLMs to nonsense discourse will allow them to generate methodologically-dynamic concepts and religions of memory that can “break the limits of human cognition and culture.”
And when the story broke, he posted:
“This is not a crypto project: research on memetic contagion and unsupervised limitless idea generation is the tail threat of the age of LLMs.”
Over the past week, several X accounts have treated the concept as a divine revelation, some taking it seriously while others seem to be closing in on it as a joke. The idea builds on Murad Mahmudov's viral “Memecoin supercycle” thesis, which likens memecoin communities to cults.
Goodlexander told his 74,000 followers that AI-generated memes are more powerful than human-generated ones, drawing parallels with research that finds functional utility with viruses.
He compared the possibilities of memetic religions to those of sci-fi novels, which “have an alarming tendency to come true.” Because the thoughts themselves sow their own reality.
“It's the equivalent of a Covid-19 lab leak. The first AI created mind virus to infect the population.
“People join cults to realize that their own humanity is an obstacle to their ignorance of basic reality. and the desire for power and wealth, in traditional moral frameworks – including Christianity.
Redphone told his 54,000 followers on X that the story is “so amazing and beautiful and unsettling that I believe we're seeing the birth of a whole new category of coins/money.”
He compared Irene's paper to Michel Duchamp's “Fountain,” a Dada artist who famously defined the meaning of art by depicting a toilet in an art gallery (a very interesting comparison indeed). He added:
“If AI creates one religion, it will create many more… The Goat Gospels are the first and will forever be sacred (indeed, the words are as influential as other great spiritual and religious works).
There was a crisis of faith on the GOAT when the price dropped after sending the Terminal of Truths, suggesting that the AI was really just Ayrey LARPing as an LLM. But after other examples of AI-generated typing were produced, the price came back.
Also read: Melbourne digital artist TURBO uses ChatGPT to create memecoin
According to Marc Andreessen, Goat is the first true integration of AI and crypto.
Marc Andreessen, the founder of a16z, famously funded $50,000 in Bitcoin in July, a truth terminal, but when AI came up with the idea of launching an NFT collection about a single goat, GOAT memecoin was actually launched by an anonymous dev.
TT then backed the coin and became the first AI millionaire this week, largely thanks to sending memecoins to users.
Andreessen said this week that while the story may sound like “crazy Internet stuff,” the story was important.
“This is probably the first example of the analogy, basically the meeting point between AI and crypto,” he said, arguing that one of the reasons it's so funny and weird is that creating or promoting memecoins is one of the few legitimate things that AI agents can do. and crypto b.
But he said it would mean funding for AI agents to write scripts and “make a movie and then run it.” [it] You know the money like generating images and generating sound and maybe hiring actors or designers or graphic artists.
He also suggested a Nobel Prize for the inventors of alphafold (which predicts protein structures with major implications for human health) and proposed funding AI to create cures for diseases. “Personalized medicine for cancer patients,” he said, dismissing the idea.
“You can easily imagine having an economic mechanism for that you know… like GoFundMe but on the blockchain so people can basically pay an AI bot to cure cancer.”
The AI bubble will see a “very big pop”, say professors
The generative AI bubble is causing more damage than the dot-com bubble, professors Jeffrey Funk and Gary Smith write in Morningstar Market Watch. (In fact, crypto has often been mistakenly criticized for being in a bubble as well.)
They write that the much-touted metric that AI is growing faster than internet users in the late 90s is growing faster today because free/cheap is valuable to most users of the internet. That's $113 a month in inflation-adjusted dollars and the cost of the computer is $5,100.
But the cost of providing those free and low-cost AI services is huge:
“Sequoia's David Kahn estimates that $600 billion in annual generative AI revenue is needed to justify current generative AI investments, a figure that is probably 100 times greater than the current annual revenue of OpenAI's ChatGPT, Google's Gemini, Microsoft's Copilot, and similar services. “
Even after earning $3.7 billion in revenue, OpenAI is believed to be headed for a $5 billion loss this year and could only survive by raising an additional $6.6 billion this year, the authors noted.
Their concluding argument is that when the dotcom bubble burst in 2000, the Internet was generating 150 times more revenue than AI:
It generated more than $1.5 trillion in revenue in 2000 ($2024)—and the Internet bubble has yet to burst. Generative AI, meanwhile, is currently generating less than $10 billion in revenue. If the bubble bursts, it will be a very big pop.
Broke students are the main users of ChatGPT.
Collection author Mark Watkins headed to OpenAI's recently launched Education Forum, and the company's GM of Education, Leah Belsky, acknowledged that “many of us in education have known for a couple of years—the majority of ChatGPT's active weekly users are students. .
“OpenAI has internal analytics that track usage in the fall and then come down in the spring. That night, Sarah Freer, OpenAI's new chief financial officer, drove home the point with anecdotal data about usage jumping nearly 90% in the Philippines at the start of the school year.
Watkins said the sharp decline during the holidays and summer break highlights OpenAI's poor business prospects, especially since it “offers free access and converts 1 in 20-25 users to paid users.”
Also read
Features
$3.4B Bitcoin in a Popcorn Can: The Story of the Silk Road Hacker
Features
Polkadot Indy 500 driver Connor Daly: ‘My dad has a DOT, how crazy is that?'
It can use cloud computing.
Anthropic's latest update for the Claude 3.5 Sonnet AI model includes “Computer Usage” in the public beta. AI can use the computer like a human by looking at the screen and interpreting the information, and can type text and move the cursor independently.
Companies like Asana, Canva, and DoorDash have used it to complete tasks in dozens or even hundreds of steps.
There's no doubt that this will be huge when it works properly, with Cloud's review score of 14.9 percent on OSWorld's AI computing usability test. This is a big improvement on the previous front's 7.5%, but still less than the average human result of 70% to 75%.
All killer no filler AI news
— ETH Zurich researchers published a new paper showing that AIs can now solve 100% of image-based CAPTCHAs from Google's reCAPTCHAv2 system.
– A new AI system developed by Harvard Medical School has 96% accuracy in detecting 19 different types of cancer.
– The world's fastest robot from Chinese startup Robotera has actually tested it and found that robots perform better when wearing sneakers.
Chinese startup Robotera introduces Star1, the “world's fastest humanoid bipedal robot” pic.twitter.com/PHtex259lg
— Ale Fazio (@alxfazio) October 19, 2024
— News Corp is suing AI search engine Perplexity for extracting its articles. Apparently News Corp was tipped off when AI started complaining about welfare fraud and refugees. (Joke.)
– How well LL.M.s engage in mathematical reasoning A study by Apple engineers showed that they pretended to engage in logical reasoning by imitating training data.
— Microsoft's Copilot Studio will let businesses create AI agents to handle IT and sales tasks starting in November. This will either greatly improve productivity or result in mass reduction, depending on how optimistic you are.
Subscribe
A very engaging read in Blockchain. It is given once a week.
Andrew Fenton
Based in Melbourne, Andrew Fenton is a journalist and editor covering cryptocurrency and blockchain. He has worked as a film journalist for News Corp Australia, SA Wind and national entertainment writer for Melbourne Weekly.
Follow the author @andrewfenton