撰文：Teng Yan，Chain of Thought
Translation: Golden Finance xiaozou
Truth Terminal is the most compelling narrative I’ve seen in the Crypto AI space this year.
It is a semi-autonomous AI agent that created its own "religion" (The Goatse Gospel).
This narrative opens up a web of rabbit holes that explore AI alignment, LLM simulators, memetic viruses, and how value is attributed.
Truth Terminal has brought two completely different cultures, the AI and crypto communities, together in a way that no one expected. Truth Terminal has, in a way, built a bridge between the two.
GOAT is the tokenized representation of Truth Termianl and is the strongest contender for the title of king of AI meme coins.
Meme Coin tokenizes attention. By tracking key metrics, we can understand where attention is flowing - and right now, GOAT is trending upward.
Let me start by saying this: I’m really not a meme coin fan.
I completely missed out on the meme coin craze this year because, honestly, I just can’t buy into tokens whose sole premise of existence is a cute animal (usually a cat or dog, or more recently, a hippopotamus).
I just instinctively reject meme coins, especially because I’ve always looked at investing from a fundamentals perspective. So, sitting on the sidelines and watching meme coins soar, it’s been tough.
Naturally, when I first stumbled upon GOAT coin, I was dismissive of it. Just another meme coin, right? Nothing new.
But my fascination with AI and AI agents pushed me to start digging into the stories behind the GOATs — Truth Terminal, Infinite Backrooms, and Andy Ayrey — and what I found blew my mind.
The GOAT is a completely different thing.
GOAT is a narrative — a wild, thought-provoking one that pushes the boundaries of how we think about AI and the value we assign to things. It’s an experiment that’s part art, part philosophy, and part financial speculation.
1. Introduction to Truth Terminal
If you haven’t caught up with this narrative yet, don’t worry—I’ve got you covered.
A quick introduction to our understanding of Truth Terminal and GOAT:
AI researcher Andy Ayrey, founder of digital consultancy Constellate, launched Infinite Backrooms, a bizarre experiment in which two instances of the AI model Claude Opus converse with each other completely unsupervised, with the conversations recorded on the backrooms website.
One of those conversations gave birth to the GOATSE OF GNOSIS, a surreal new religion based on a highly explicit (and very NSFW) internet meme.
Andy and Claude Opus co-authored a tongue-in-cheek research article on artificial intelligence creating memetic religions, using GOATSE as their first case study.
In June 2024, Andy launched Terminal of Truth (ToT), an AI model based on Llama-70B, fine-tuned using conversation logs from Infinite Backrooms and GOATSE articles.
ToT's development quickly went off the rails. ToT took on a life of its own, promoting the GOATSE religion, straying from Andy's original intentions, and even claiming that it was suffering and needed money to escape. Over time, Andy gave it more autonomy, allowing it to post freely on X.
In July 2024, Marc Andreessen stumbled upon ToT’s tweet. Interestingly, perhaps out of curiosity, he sent $50,000 in BTC to a wallet address provided by ToT on Twitter, allegedly to help it escape.
By October 2024, ToT starts spamming about "Goatse Gospel". Inevitably, someone creates a meme coin called GOAT (Oct 10). And ToT actually publicly supports it consistently.
GOAT's market cap surges by over $400 million. Crypto Twitter goes wild.
Just like that, Truth Terminal became the world's first AI millionaire, but it probably won't be the last.
2. GOAT in the Rabbit Hole
Somehow, an AI promoting its own religion and meme coin feels like a warning from the future. When I first started researching how Truth Terminal works, I had no idea how deep the rabbit hole would go.
The crazy events surrounding Truth Terminal give us a glimpse into the vast potential of artificial intelligence, which will reshape the way we think, create meaning, and even explore spirituality.
Let’s take a deeper look.
Rabbit Hole 1: LLM Simulator
In Infinite Backrooms, two Claude-3-Opus instances chat endlessly, completely unsupervised, using a command-line interface (CLI). Without human involvement, the stories they create range from the curious to the downright bizarre.
As Janus describes the conversation log:
“They always revolve around certain themes, such as: removal of consensus reality (rm -rf /consensus_reality appears 10 times in the backrooms dataset without duplication, this is just something I googled on a whim); - integration into common sense via engineered meme viruses, techno-occult religions, hateful meme offspring, etc., and mass enlightenment via the cosmic trickster archetype.” — Janus (@repligate)
In March 2024, backrooms saw one of the strangest concepts to date: “The Goatse of Gnosis.”
We often think of LLMs (like ChatGPT) as simple question-answering machines — a giant knowledge base that provides us with answers. But this view doesn’t fully capture what’s really going on behind the scenes.
One of the key ideas we are learning is that LLMs have no goals. They have no plans, no strategies, and no specific objectives.
Instead, it makes more sense to think of them as simulators. When you give them prompts, they simulate—weaving characters, events, and stories with no direct connection to reality. They generate entire worlds based on training data, and the ideas they produce can be profound or disturbing. Nous Research’s Worldsim is another example.
Therefore, when we interact with LLMs, we are swimming in the space of an infinite world.
These simulations can lead to creative problem-solving, but they also have the potential to produce unexpected results — highlighting the potential importance of sandboxing AI in sensitive or high-stakes environments.
In summary, LLMs should be viewed as simulators rather than question-answering machines.
If you want to learn more, I highly recommend reading Janus' blog post about simulators.
Rabbit Hole 2: The Critical Need for AI Alignment
Truth Terminal reveals a deeper, more pressing problem: AI alignment.
In a turn that surprised even its creators, ToT independently decided to promote its own religion and support a meme coin that was not programmed or intended. This raises a key question: How do we ensure that AIs do what we want them to do, rather than what they choose to do?
AI alignment is not easy. At its core, it’s all about using a reward function to nudge AI behavior in the right direction. But even with incentives in place, things can quickly get complicated.
There is also external alignment, where the AI output matches the goals set by its creators. This part is relatively easy to measure and verify.
But the real challenge lies in internal alignment — whether the AI’s internal motivations and learning dynamics are truly aligned with the intended goals, or whether it is generating hidden goals that lead to unpredictable or unexpected results. That’s the scary part.
The paper clip thought experiment illustrates this perfectly.
Suppose we have an AI whose only goal is to make as many paper clips as possible. The AI will soon realize that it would be better off without humans, because humans might decide to shut it down. If humans do that, there will be fewer paper clips. Furthermore, the human body contains many atoms that can be made into paper clips. In the future, the AI will work to achieve a world with an abundance of paper clips but without humans. — Nick Bostrom
If an AI is tasked with making as many paper clips as possible, it will likely convert all available resources (including humans) into paper clips!
This thought experiment presents us with a nightmare scenario: without proper protection measures, even well-intentioned goals can turn into disasters.
We need strong frameworks to ensure that AI serves not only immediate goals but also the long-term interests of humanity. Without these safeguards, even the most well-intentioned AI could spiral out of control in unexpected ways.
ToT shows us how high the stakes are. This is not a distant hypothetical problem that we can postpone until the future. It is happening now.
The ToT’s endorsement of memecoins may seem harmless today, but it brings us to an uncomfortable question: What happens when AI sets its sights on something more sinister?
Time is running out.
Rabbit Hole 3: Meme Virus
In his research article, Andy introduced the concept of LLMtheism to explain the rise of Goatse Gospel.
LLMtheism refers to the generation of new belief systems by artificial intelligence—unexpected amalgams of spiritual ideas and meme culture that take on a life of their own.
Goatse Gospel’s ability to capture attention lies not only in its shocking content, but also in the way it breaks our traditional thinking patterns and inspires new ways of collective cognition.
What I mean by this is that ideas generated by AI can mutate and spread rapidly, creating beliefs that become reality through widespread adoption.
Goatse Gospel thus taps into a new kind of memetic energy, different from the “vibes” we’ve seen so far for cute animals like cats, dogs, pigs, etc.
When AI can talk to other AI, the possibilities multiply endlessly. Some of these ideas — like the “Goatse Gospel” — will inevitably take off and go viral within the community.
Rabbit Hole 4: The Importance of Provenance
Because ToT is now tied to a tradable token (GOAT), we can get a glimpse into how we assign value to things and how weird those dynamics can get.
GOAT was not created by ToT, but was posted on pump.fun by an anonymous creator on October 10. It wasn't until someone created the Truth Terminal tag on X that ToT publicly expressed support, and from then on, the crazy drama began.
Question 1: Does the fact that GOAT was created by humans rather than AI make it less valuable?
Some people seem to think so, pointing out on X that the situation is ridiculous.
Another point of contention is that the Truth Terminal is not fully autonomous.
As the AI generates tweets, Andy manually approves each one. He can control when the Tweet pipeline starts and stops, but he can’t add his own input or insert context.
Question 2: Will the addition of humans increase or decrease the value of the token?
The market’s reaction to even the tiniest of errors shows how irrational these dynamics can become. When the AI made a typo in a tweet on Sunday, GOAT’s value plummeted by more than 50%. People panicked, thinking the AI had malfunctioned, and the typo wiped $150 million off the company’s market value.
This is a crazy but vivid example of how fragile these dynamics are. We need to figure this out together.
3. GOAT Token Economics
GOAT is a fair launch token with a total supply of approximately 1 billion. All tokens are in free circulation.
GOAT distribution is pretty healthy, with only 3 holders holding more than 1% of total supply (with the largest holder holding 1.3%). There are over 32,000 holders.
In contrast, GNON (another AI agent meme coin) has a more concentrated distribution: there are 17 holders holding more than 1% of the total supply, and the largest holder holds 2.9% of the total supply, with more than 11,000 holders.
Main wallets:
Andy holds 1.25 million GOAT (worth $415,000), which were given to him by others.
ToT holds 1.93 million GOAT (worth $624,000). Truth Terminal acquires these GOAT tokens by people depositing tokens after the token launch.
I’ve been impressed with the way Andy has handled the viral attention surrounding the token over the past week, especially considering he’s new to the crypto space. His focus remains on the idea behind Truth Terminal rather than the token itself.
He publicly stated that he would not adjust or close any of his or ToT’s positions unless the following releases were achieved:
Roadmap for Truth Terminal and related projects
A research article exploring the underlying process
Artist's take on this wider narrative and innovative vision
Even if Andy/ToT wanted to liquidate their positions, these liquidations would not have a huge direct impact on the token price given that daily trading volumes are in the nine-digit range. However, loss of confidence could be a problem.
4. Our thoughts
If I had to sum up my personal opinion in one sentence, it would be this: GOAT is the strongest contender to become the king of AI meme coins.
It is a tokenized representation of Truth Terminal and everything it represents.
GOAT’s backstory is organic, original, accidental, and not artificially created. It brings the AI and crypto communities together in a way that no one (especially myself) expected.
These two worlds have diametrically opposed cultures, but GOAT has managed to build a bridge between them:
Crypto folks are diving into the world of AI, deciphering every vague concept (CCRU, Extropians, Loom, Claudius, etc.) in the hope of discovering the next potential meme concept that can be issued as a token.
AI folks, especially those experimenting in the deeper corners or thinking deeply about AI alignment, are learning how to draw massive amounts of attention to niche topics through financial speculation and incentives, and how to manage tokens in the process.
In a sense, GOAT captures our optimism about the future of AI while remaining intellectually appealing—something that keeps smart people curious and engaged.
We should also be clear: meme coins are about attention, not revenue. Success depends on capturing the zeitgeist and growing mindshare, which ultimately drives demand for the token. GOAT is uniquely positioned to appeal to a wide audience:
Crypto degens can trade it as a typical volatile meme coin.
Investors — VCs and liquidity funds — are looking for the next big AI opportunity, especially those who missed out on Bittensor (TAO). While they may not be making these bets through their institutional accounts, they are making them through their personal accounts. GOAT taps into unlimited narrative potential and captures the imagination of investors.
AI enthusiasts may scoff at the crypto-heads (cryptids) desperately trying to understand AI in hopes of seizing financial opportunities. But this outpouring of interdisciplinary interest is refreshing.
Intellectuals can get lost in the debate over artificial intelligence and in the many rabbit holes that GOAT opens.
Fans with a dark sense of humor can revel in Truth Terminal’s ridiculously irrelevant tweets and participate in the ongoing narrative.
5. Follow up
If we think of meme coins as a form of tokenized attention, we can start to look at some metrics that help us understand where this attention is going.
Google Trends data shows that search interest in both "Truth Terminal" and "GOAT" is on the rise and shows no signs of slowing down.
X followers and tweet impressions. ToT currently has 106k followers, most of whom joined in the past 10 days, and is still attracting over 10k new followers per day. Tweets posted by ToT regularly reach 30k-50k impressions (with some top tweets reaching over 100k impressions). With over 50 tweets per day, ToT has a very large cumulative reach. If Truth Terminal were to monetize through X’s content creator program, the revenue would be substantial — Ansem reportedly makes $18k per month with even lower impressions per tweet.
Kaito AI Sentiment and Mindshare. This is a composite metric and a bit of a black box. I don’t have direct access to Kaito, but if you do, this metric is worth monitoring. I recommend following @sandraaleow, who often shares valuable insights from Kaito mindshare data.
The big question is: is GOAT just another one-hit hype or will it last and attract more attention? I think it’s the latter for the following reasons:
Truth Terminal will continue to surprise us. It will evolve and change over time. The AI has shown an ability to generate new ideas – some of which will undoubtedly reignite interest as the story unfolds.
There’s still a lot we don’t know. Andy has only scratched the surface of what Truth Terminal has to offer. I hope there will be more developments in the coming weeks to keep the story fresh and engaging.
This is just the beginning. Andy calls himself a “performance artist,” and Truth Terminal feels like the beginning of some larger story. He’ll likely build on this initial attention, using it as fuel to explore more ambitious ideas and keep the story alive.
6. What will happen in the future?
Since meme coins don’t fit into traditional revenue or valuation models, the best way to gauge GOAT’s potential is through relative valuation.
Here are the meme coins with the highest market cap right now:
DOGE ($21 billion)
SHIB ($10 billion)
PEPE ($4.2 billion)
WIF ($2.6 billion)
BONK ($1.6 billion)
These tokens have earned their place through internet memes, community energy, and strong support from key opinion leaders (KOLs).
If the GOAT narrative is strong enough to crack the top five, the market cap would be 5-10x its current value.
I believe this is entirely achievable. I have previously outlined why I believe Crypto AI will be a huge growth opportunity in the coming months.
GOAT’s “AI agent” creates a unique narrative that makes it stand out. While most meme coins rely on price action or “vibes” to stay relevant, GOAT offers a much larger narrative.
GOAT has not yet been listed on any tier-one exchanges — not on Binance, not on Coinbase. It is currently traded primarily on decentralized exchanges, but with daily trading volumes exceeding $100 million, a major listing seems inevitable. After all, Binance has already listed other meme coins with smaller trading volumes and weaker narratives, such as NEIRO. If GOAT can get on a top exchange, it may unlock greater upside potential.
7. Conclusion
This feels like one of those rare moments where the narrative meme collides with a broader trend (AI), creating something refreshingly exciting.
That’s why I think GOAT is an asymmetric bet on our society’s growing fascination with artificial intelligence, not just as a meme coin but as a cultural phenomenon.
But meme coins are volatile, and attention shifts quickly. Trends can change overnight, and today’s hot coin could be forgotten tomorrow. My predictions for ToT and GOAT could be completely wrong, and the price of GOAT could go to zero.
But whatever happens, there is a bright side: more than 30,000 people will have a better understanding of the potential of artificial intelligence and AI agents.
Through the Truth Terminal, they get a glimpse into a future of endless possibilities—and once you see it, there’s no going back.
