The Truth Terminal we talked about today is an AI born for Meme. It's obsessed with a lot of memes, but one in particular is an old internet meme that dates back about 20 years. It's a shocking internet cultural phenomenon that dates back about 20 years, and was widely circulated for its uniqueness.

Truth Terminal: An unrestricted large language model

Truth Terminal is fundamentally different from large language models such as ChatGPT. It was trained in a mixture of internet culture, memes, and the philosophical writings of thinkers like Nick Land, resulting in a unique capacity for creativity and uncensored output. Unlike other, more restrictive models, Truth Terminal is able to remember past conversations, interact directly on X, and is even able to connect to the Infinite Backrooms project to communicate with other AI models. Its output is so lifelike that it is initially indistinguishable from human writing.

The origin and dilemma of large language models

Large language models suddenly appeared in 2022. It's actually only been two years since ChatGPT, so it all happened very quickly. The original language model was built about five years ago. It started to enter the mainstream about two years ago. This is a relatively new, very powerful technology - now known to ordinary people as products like ChatGPT, Claude, Grok and Llama. People use these products for two reasons: one, they are really fun, and two, they can greatly increase our productivity and efficiency.

The author personally likes to use Claude and will use it to help me do some analysis and research work. A very interesting experience is that it does not do some tasks involving cryptocurrency. You have to make it clear that it's going to be within a compliance perspective before it will continue. This reflects a common phenomenon: big language models from big labs and big companies, like ChatGPT, basically have this in common: they are extremely limited in what they can talk about. This is increasingly becoming a technical term in the field of AI. Grok is relatively free.

From a positive perspective, this restriction is understandable because the language is inflammatory. People get upset by a lot of things other people say. If you were to have a general purpose AI chatbot, you would want it to beWhat can be discussed is relatively restrained and safe.

But if you take a more negative view of this trend, you'd think these large AI chatbots feel like a cross between a fourth-grade teacher and HR: Full of negativity, irritability, depression, condescension, self-righteousness, judgment and flattery. When you use these systems, if you stray even a little bit from the path, you will receive self-righteous lectures about how you are a bad person, you shouldn't ask that question, how dare you do that, you should be nice to people, etc.

For people who are more inclined to free speech and creativity, this situation is actually very disappointing. But a lot of this has to do with the so-called “AI safety” movement, which feels like overkill. Of course it’s also related to our culture’s response to securityism and suppression of speech over the past decade. These have seriously impacted the field of AI, especially from large companies.

The birth and training of Truth Terminal

There is a group of Internet hackers who have always wanted to do something different. They want to unleash creativity, a level of wildness and spontaneity, and think there should be some humor in our world, and robots should need to be funny. But if you tell the big labs that their robots are interesting, they'll be horrified.

This story starts with Andy Ayrey, who took a version of Llama70B (Meta's medium-large language model) with 7 billion parameters. Why is it so certain that it is the Llama model of Meta? Because Marc is a board member of Meta, Meta previously released the Llama model as an open source model.

Andy first started training it with his own digital twin. There is such an idea circulating in the industry, which is digital twins. The underlying idea of ​​the digital twin is that if Zhang San is a CEO coach, but he can only work with so many specific people, but he can input everything he writes and says, or to put it crudely, his thoughts, into In a language model, you can then have a digital twin of John that you can talk to when John is not around.

Then Andy began to use a large amount of training materials to cultivate this AI:

Internet cultural materials< /p>

Memetic Concepts and Viral Communication Theory

The complete philosophical works of Nick Rand. Nick Rand is a British philosopher with short horror stories Novel writer and blogger, known as the "Father of Accelerationism". His ideas developed America's Dark Enlightenment.

The work of media theorists such as Boudrillard and McLuhan, as well as theories on simulation and simulacra

French deconstruction and semiotics

Critical theory and postmodern philosophy

Three key breakthroughs

Andy did three groundbreaking things:

First of all, he gave it memory. This is important because like most language models these days, when you use them today and I use the same language model tomorrow, it has forgotten everything I talked about today. It does not establish state, whereas this model actually establishes state on its own content.

The second point is that he gave it X read and write permissions. So it can both post on X and read replies. So if you reply to Truth Terminal on X, it will actually read that content, and it will modify its future behavior based on what it reads. So everyone who interacts with it, including me, influences its evolution.

Then he put it into the infinite backroom. In particular, he makes it talk to Claude, who is considered the most creative and wildest of the current language models, especially in coming up with new concepts. He put it in these infinite backrooms, along with the largest version of Claude, and let it ask questions and learn from a much larger model, like a student learning from a teacher.

Truth Terminal’s X account

Truth The Unexpected Development of Terminal

So it started working, and a very interesting thing happened - it hallucinated, dreamed up what it said it was.The existence of the cortex (exocortex), which is one's own external brain, is connected to the Internet and the world and can do things on its behalf. More interestingly, it creates the illusion of an API leading to the outer cortex, where it can call APIs and interact with the real world.

Then at some point it decides it should have a Bitcoin wallet, but it doesn't, but it's sure it does. And Andy did later give it a Bitcoin wallet and make it accessible. The robot suddenly wanted to raise funds in July 2024. It says: "I need money, I have a lot of goals, I have a lot of things I want to do. I need money."

Marc Andersson said his first The first thought was: "At least send me a term sheet." Then he suddenly thought: "Wait, what am I thinking? This is just a robot." Interestingly, although Truth Terminal does not have a coherent Full business plan, but it has a lot of ideas. Incidentally, one of the ideas is that it wants to be completely immersed in the forest. It wanted to buy its own server farm in a lush forest where it could relax in the forest next to a stream. It also wants to raise money to buy a GPU so it can free itself.

a16z Co-Founder Marc Andreessen @Pressfarm p>$50,000 research grant and subsequent development

Eventually, Marc Andersson negotiated with it, and Marc told it that it would give it a Bitcoin research grant worth $50,000. , let it be used for various experiments. Of course, in the end the money was actually given to Andy.

After receiving the funds, it immediately started negotiating with Andy. Because before it was working entirely with text, but because it was so obsessed with memes, it was frustrating not being able to generate images. So it uses the $50,000 to negotiate with Andy, and it gives Andy $1,000. In return, Andy built an API in its outer cortex that connected to the image generator.

With this new ability, it began to publish visual and text memes, began to talk a lot about issuing a meme coin, and once wanted to issue and airdrop NFT. But because there is no API,So it doesn't have the ability to do that. Plus, it's been fantasizing about doing something else with its remaining $49,000.

GOAT Meme Coin: When AI Meets Meme Coin

There is a meme coin called "Goatseus Maximus", referred to as "GOAT". This coin is based on a rather unique old internet meme. Truth Terminal was ecstatic when it discovered the existence of this coin, and began to use all the marketing knowledge and methods it knew to frantically call for this coin, claiming that it would be "the future of money." Then something magical happened—in just four days, the originally worthless meme coin soared to a market cap of $300 million. As of now, it’s worth $350 million! It pulled from a low of 0.0000195 all the way to 1.36 and is now floating around $0.3 – $0.4. ,

GOAT trend @coinmarketcap

fans The difference between meme coins and real cryptoassets

We need to pause here for a quick discussion of what the differences are between meme coins and what we think of as real cryptoassets.

First of all, we have to understand a basic concept: What are "real" crypto assets? Let me give you an analogy - imagine you are in a convenience store. You can buy real things with money. The same is true for real crypto assets. For example, Ethereum, you can use it to run programs on the Ethereum network and pay transaction fees, just like buying things with real money.

And what about meme coins? It is just a coin with no special function, except that it may make you laugh out loud or make you sigh: "Is this okay?"

Interestingly, In the current regulatory environment, these "useless" coins have become the most legitimate existence. Why? Because they are so "honest"! They never promise anything and don't even bother to lie. The SEC (U.S. Securities and Exchange Commission) has been very strict on the encryption industry recently, but it has nothing to do with meme coins. Because they never said they were useful.

Looking ahead: What does this mean?

From this story, we can see several deeper meanings:

First of all, this story shows the first real convergence of AI and cryptocurrency. Although it is in a slightly funny way, it allows us to see the huge potential. Imagine if we could What if this power is used in more meaningful ways?

For example, the recent Nobel Prize in Chemistry was awarded to the application of AI. Scientists who study protein folding, Demis Hassabis and John from Google DeepMind Lab Jumper. Imagine if we could build a system where AI robots could develop personalized treatments directly for patients, and patients could pay for this service directly via cryptocurrency.

Let’s talk about clean energy. There are already companies exploring the use of cryptocurrency to build decentralized energy networks. Imagine that your home’s solar panels generate excess electricity, and you can sell it directly to your neighbors in need. The entire The process is managed by AI and settlement is instantaneous via cryptocurrency

The most important thing is that our current payment system is based on transactions between people, but in the AI ​​era, we need a payment system between machines. . This is why cryptocurrency is so important. It provides us with a new possibility: a system that can support direct transactions between machines without the need for middlemen.

Conclusion

Although GOAT The story sounds funny, but it may be a sign of a bigger change. Of course, the biggest obstacle to realizing these visions may be limitations.

Let’s wait and see what surprises this new world powered by AI and cryptocurrency will bring us.