The past year has brought some surprising realizations about AI, productivity, and meaning. I've discovered that while AI can dramatically accelerate initial progress on projects—perhaps getting you 80% of the way there in a fraction of the usual time—the final 20% often remains stubbornly time-consuming. This "AI productivity paradox" raises important questions about the value of AI-accelerated work.

The "AI value paradox" is more troubling: as AI makes some outputs easier to produce, their market value tends to decrease. For example, how valuable are these blog posts when similar insights could be gained through a direct conversation with an AI system? When everyone can easily create high-quality content, what makes content truly valuable?

These questions led me to spend more time offline, finding spaces away from the constant hum of technology. Ironically, this distance has renewed my excitement about AI's potential, but in more personal ways. I’ve been writing by hand more - there's something about the physical act of writing that AI can't replicate. But I've also been using AI to generate books that I then print and bind myself. There's also something satisfying about combining AI's creative capabilities with the tactile craft of creating physical objects that can sit on my shelf. While AI can write the content, it can't bind the pages or create that feeling of holding a finished book in your hands (yet, anyhow).

This seemingly backward approach—using AI to create content that I turn into physical books—has helped me understand something important about finding value in an AI-accelerated world. It's not about resisting AI's capabilities or trying to compete with them. Instead, it's about finding ways to use AI to enhance rather than replace the activities that bring us personal satisfaction.

I've also realized that I don't need to approach every AI project with the goal of scaling it into a business or sharing it with the world. Sometimes, the most rewarding projects are the ones we create just for ourselves. This feels particularly relevant as AI makes producing and publishing content increasingly easy. Just because we can easily create and share something doesn't mean we should.

This perspective shift has made me more excited about AI's potential, not less. When I stopped trying to keep pace with technological change and started focusing on what brings me genuine satisfaction, I found more meaningful ways to integrate AI into my work and creative projects. The pressure to constantly produce and publish began to fade, replaced by a more thoughtful approach to creation.

That's partly why I've been quiet on this blog. While I could easily use AI to generate regular posts about AI business models and autonomous systems, I've been questioning whether adding more content to the already overwhelming flood of information serves any real purpose. Sometimes, stepping back and remaining quiet is more valuable than contributing to the noise.

This isn't goodbye—I'm leaving the door open for future posts when I have something meaningful to share. For now, however, I'm enjoying using AI more personally, creating books for my bookshelf rather than an online store, and working on various other projects. Maybe someday, I'll share pictures of my growing library or other offline projects I use AI for. But for now, there's something satisfying about keeping some things just for myself.

For those following the Dabble.AI project, thank you for joining me on this journey. What started as an exploration of autonomous business models has evolved into something more personal - a search for authentic ways to work with AI that enhance rather than replace the activities that bring us joy.

1. What product or service offerings could a single-person company scale to 1B? 

2. What are the systems and technologies that make it possible?

Generally, the answer to the first question is likely something virtual or digital. However, the market is also a significant consideration. You’d need a market with the dynamics and size to enable and support a 1B+ single-person company. Let’s consider ebooks as the product. How likely would a single-person ebook publishing company be able to scale to 1B+ in annual revenues? Based on today’s ebook market, I’d guess it's not very likely. But what if AI changes the market for ebooks?

When I think about ebooks today, I think about books sold to humans. But what if AI agents are also consumers of ebooks in the future? Sound crazy? That market is already starting to emerge. The “ebooks” are vector databases, systems prompts, and retrieval augmented generation (RAG) systems. They provide AI systems with information to expand and enhance their current base of knowledge—just like ebooks can for humans. Is that an ebook? That’s not the point. The point is that new markets are emerging, and existing markets are changing. So, we can’t just think about products and markets based on how we define them today. 

Let’s envision a publishing company that sells “books” to AI agents. I’m not talking about AI writing books for humans but about books written for AI consumption. What could that look like? What technologies could make it happen? Incumbent distribution channels like Amazon.com focus on human consumers, so someone must create a new distribution system. This system would need to be discoverable by AI agents. The agents would also need a way to evaluate the value of the books and, of course, buy them. The distribution systems would need to ensure that the books don’t get “plagiarized” with a mechanism to prevent AI agents from including the “books” in their internal language models. 

I’m not exactly sure how it would all come together. It would likely involve an AI agent discovery system or protocol that agents could use to discover other AI agents based on their capabilities. It could also leverage blockchain technology to enable transaction processing, NFTs, and immutable on-chain access logging to prevent “AI plagiarism.” Perhaps there is also a blockchain-based rating system that lets agents provide a feedback loop to grade and help improve the usefulness of individual books. 

Imagine this exists, and you have a best-selling book when AI agents vastly outnumber the human population. What if billions of agents buy fractional access to your book’s NFT or pay each time they access it? In that hypothetical scenario, could selling ebooks be a billion-dollar opportunity for a solopreneur?

That’s it for today. But if this post inspires you to start building a company that provides “books” for AI systems, or if you’ve already started, let’s talk—I’m also an angel investor.

In part 2, I discussed how blockchain technologies might enable multi-agent AI systems across organizational and platform boundaries. In this post, I'll provide some examples of how specific blockchains could play into that. Specifically, I'll discuss Ethereum, Chainlink, IPFS, Hyperledger Fabric, Ocean Protocol, Polkadot, and Golem.

I won't get into technical details. I'm also assuming you're willing to read up on the blockchains I'm using as examples. Or that you're familiar with them already. With that said, let's dive in.

1. Ethereum Smart Contracts: Ethereum is the most widely used platform for smart contracts, which are self-executing coded contracts with the terms directly written into code. AI agents could use Ethereum smart contracts as a coordination mechanism, defining the rules of engagement, data sharing, and reward distribution between agents. Smart contracts could also be used to create decentralized marketplaces where AI agents can offer services or request tasks to be completed by other agents.

2. Chainlink Oracles: Chainlink provides decentralized oracle networks that can feed real-world data into smart contracts on various blockchains. AI agents could utilize Chainlink oracles to access trusted off-chain data sources, APIs, and even AI/ML models. This could enable AI agents to incorporate real-time data into their decision-making and interact with external systems through smart contracts. Chainlink's reputation system for oracles could help AI agents select reliable data sources.

3. IPFS (InterPlanetary File System): IPFS is a decentralized protocol and network for storing and sharing data in a distributed file system. AI agents could use IPFS to store and share large datasets, models, or computation results in a decentralized manner. IPFS content addressing could be used in smart contracts to ensure data integrity and provenance.

4. Hyperledger Fabric Channels: Hyperledger Fabric is a permission-based blockchain platform that supports private data collections through its channel architecture. AI agents could utilize private channels to securely share sensitive data or models with specific authorized parties, enabling collaborative learning while maintaining data confidentiality. This could facilitate secure multi-party computation and federated learning between AI agents from different organizations.

5. Ocean Protocol: Ocean Protocol is a decentralized data exchange platform that allows data owners to sell access to their data while maintaining control and privacy. AI agents could use Ocean Protocol to securely access and purchase datasets for training or analysis, enabling decentralized data marketplaces. Ocean's data tokens and compute-to-data capabilities could enable AI agents to perform computations on data without direct access, preserving privacy.

6. Polkadot Parachains: Polkadot is a multi-chain network that allows different blockchains to interoperate and share security. AI agents operating on different parachains could use Polkadot's cross-chain communication to coordinate and exchange data or value. Polkadot's shared security model could provide a trusted environment for AI agents to interact across chain boundaries.

7. Golem Network: Golem is a decentralized computing power marketplace where users can rent out their idle computing resources. AI agents could use Golem to access decentralized computing power for resource-intensive tasks like model training or large-scale simulations. This could enable AI agents to scale their computational capabilities on demand without centralized infrastructure.

These are just a few examples. There are many others, and new possibilities are rapidly emerging. The key themes are using blockchain features for secure coordination, data sharing, decentralized marketplaces, and access to decentralized resources. Now let’s consider a hypothetical but real-world example.

A real-world example

Imagine a future where a patient's health data isn't siloed in a single hospital's database, but instead is securely shared across a decentralized network of healthcare providers, research institutions, and AI service providers. Each entity operates an AI agent that specializes in a particular aspect of healthcare: diagnosis, treatment recommendation, drug discovery, etc.

These AI agents can engage in secure, auditable collaboration via smart contracts on a blockchain platform like Ethereum. When a patient grants consent, their anonymized health data is made available to the network via a secure data-sharing protocol like Ocean Protocol. The AI agents can then purchase access to this data using Ocean's data tokens without the data ever leaving the patient's control.

With access to a diverse set of patient data, AI agents can perform federated learning to improve their models while preserving data privacy. They can use Chainlink oracles to incorporate real-world data on treatment outcomes and drug efficacy into their learning process. The improved models are then made available to the network, with their provenance and performance recorded on the blockchain for accountability.

When a new patient comes in with a complex set of symptoms, their healthcare provider submits a request to the network via a smart contract. The AI agents then bid to provide their diagnostic and treatment recommendation services based on their specialized expertise. The smart contract aggregates the responses, weights them based on each agent's proven performance, and returns a personalized treatment plan to the healthcare provider.

Throughout this process, Hyperledger Fabric channels are used to maintain patient privacy and comply with regulations like HIPAA. Payments for AI services are automatically handled via the smart contract, with a portion of the fees being allocated to compensate patients for sharing their data.

In this use case, the decentralized nature of the system prevents any single entity from monopolizing patient data or AI capabilities. The blockchain provides a trusted, transparent coordination layer for the AI agents to collaborate and compete, aligning incentives around patient outcomes. The combination of secure data sharing, federated learning, and decentralized governance enables the emergence of a powerful collective intelligence for healthcare.

This is just one example, but it illustrates how the intersection of multi-agent AI and blockchain could transform entire industries. By enabling secure, decentralized collaboration between specialized AI agents, these technologies could unlock new frontiers in personalized medicine, predictive maintenance, financial modeling, and beyond.

This is just a hypothetical example. But hopefully, it illustrates how the unique features of blockchain - decentralized trust, secure data sharing, and automated coordination can amplify the power of multi-agent AI systems in ways that are difficult to achieve with centralized architectures.

This is the final post for this series. But I’m sure I’ll be revisiting the topic of AI/blockchain convergence as I dive deeper into how to develop autonomous business models.

In the last post, I discussed multi-agent AI systems and why I see them as a key enabler for significantly more capable AI systems and potentially AGI. In this post, I'll discuss how blockchain technologies will likely play a critical role in enabling multi-agent AI systems across organizational and platform boundaries.

But first a disclaimer: I'm speculating here. I'm not placing any concrete bets at this point. Things are moving fast and there are a lot of unknowns about how these technologies will evolve and interact. But I can't help imagining the possibilities. So, that's what this is - imagining possibilities.

Blockchain and Multi-Agent AI: Synergies and Potential Use Cases

There are many possibilities so I'm just touching on a few that seem like the most obvious use cases. They are:

Decentralized Coordination: One of the key advantages of blockchain technology is enabling decentralized coordination and agreement between parties who don't necessarily trust each other. This aligns well with the concept of multi-agent AI systems that need to coordinate and collaborate across organizational boundaries. Smart contracts on a blockchain could provide a secure, transparent, and immutable way for AI agents to negotiate, establish agreements, and execute shared workflows even if they are running on systems controlled by different entities.

Collective Decision-Making: Building on the previous point, DAOs (Decentralized Autonomous Organizations) on a blockchain utilize smart contracts to enable decentralized governance and decision-making. In a multi-agent AI context, smart contracts could similarly be used to facilitate collective decision-making between AI agents without centralized control. The rules and voting mechanisms would be codified in the smart contract. This could allow large numbers of agents to efficiently coordinate and converge on decisions.

Marketplace Infrastructure: Blockchain platforms with smart contract capability essentially provide a decentralized marketplace infrastructure. This could be leveraged to create marketplaces for AI agents and AI services. For example, a smart contract could allow people or organizations to post machine learning tasks, datasets, or problems, along with bounties/payments. AI agents could then bid on and complete the tasks, with the smart contract automatically validating results and disbursing payment. Think Mechanical Turk for AI agents. The blockchain provides the trusted infrastructure for this without a centralized platform.

Federated Learning: Federated learning is an approach to machine learning where the model is trained across decentralized edge devices or servers holding local data samples, without exchanging the data samples. This can help address data privacy and data siloing challenges. Blockchains could provide a shared layer for coordination between federated learning nodes run by different organizations. Smart contracts could be used to orchestrate the federated learning workflow transparently.

Beneficial AI Alignment: AI Alignment efforts are focused on the future ability to steer AI systems toward a person's or group's intended goals, preferences, and ethical principles. To that end, blockchains could be used to create a decentralized, transparent registry of AI systems. Smart contracts could specify and enforce certain constraints or behaviors of AI agents - almost like a constitution. There could be mechanisms for humans to monitor, audit, and vote on changes to AI systems via the blockchain. Maintaining decentralized oversight of powerful AI, rather than letting it be controlled by a single entity will be the goal.

Again, this is all speculative musings on my part. But from a technology standpoint - the potential is there for sure. Still, there are tons of open questions and challenges that need to be addressed. For example: blockchains can be inefficient and hard to scale compared to centralized systems. Many blockchains have limited privacy. Most AI today is centralized in large tech companies and governments which will likely resist decentralization. And there are still numerous unsolved challenges in multi-agent coordination/orchestration. Even still, the potential is exciting.

In part 3, I’ll provide examples of how specific blockchains could potentially support multi-agent AI systems. Specifically, I'll discuss Ethereum, Chainlink, IPFS, Hyperledger Fabric, Ocean Protocol, Polkadot, and Golem. I'll also present a hypothetical but real-world example to illustrate how it might all come together. So stay tuned!

This will be a multi-part post because there’s a lot I want to unpack. I'm starting with AI orchestration. I'll discuss the intersection of AI and blockchain technologies in a future post.

Let’s begin with the end in mind - a future when highly autonomous organizations are commonplace. For the sake of this discussion - I’m not talking about solopreneurs running AI lifestyle companies. I’m talking about complex organizations that today would require hundreds or thousands of employees, across multiple departments and divisions, with many millions or billions of dollars in revenue. To see that happen, we’ll need more advanced AI. So let’s start there.

The two paths to more advanced AI

In the simplest sense, there are two paths to more advanced AI - scaling up and scaling out. Scaling up is about making a single AI model more and more capable. This is the current trend (generally) with the push for bigger and bigger LLMs - large language models. However, hardware constraints are a major obstacle to continuing on that path. Just look at NVIDIA's stock performance over the past few years. That illustrates part of the problem - costs being driven up by supply and demand. But at some point, it goes beyond simple economics and becomes a matter of physics.

The short version of the physics constraint is that microprocessors generate heat and everything melts at some point. Advances in quantum, carbon nanotube, and other new computing technologies might help overcome current limitations. But scaling out is the natural evolutionary path in nature and I suspect it will be for AI also, even if we see significant hardware improvements. This is why I think AI orchestration technologies are so important. They are likely the key to significantly more advanced AI systems and potentially the enabler of AGI - artificial general intelligence.

One AI model/provider will never be enough

Even if we could scale up, a single AI model or service provider will never be enough. No single model or provider will do everything better than all the other options. Plus, the best AI for a given task will change over time, as existing systems improve, or new options become available. Also, costs, performance, data privacy, IP protection, and other business-specific considerations will drive decisions to use multiple models, providers, and tools. But whatever the reasons, as the demand for more advanced AI applications increases, staying competitive will require using multiple AI models and tools. That’s where AI orchestration comes in.

So what is AI orchestration?

AI orchestration combines and coordinates multiple AI models, tools, systems, workers (AI and human), and other related resources to accomplish goals that no single AI system or group of humans could accomplish independently. Said another way - it’s about organizing and managing AI and human resources to survive and stay competitive.

Terminology and key concepts

The terminology isn’t standardized, but the key concepts are easily understood. The best way to wrap your head around the general concept is to consider a traditional company with just human workers. In a traditional company, you have employees with different roles and responsibilities. Those employees (or contractors) use their skills, experience, tools, and other resources to accomplish tasks that align with the company's goals and objectives.

AI orchestration is similar. Just swap roles for employees and contractors with roles for AI workers/agents. And just like in a traditional company, an AI agent might be specialized in a particular area or a generalist. And again like in a traditional company, AI workers will have varying levels of skill, specialization, costs, responsibility, and decision-making autonomy.

An AI agent is usually more than a single AI model or tool. It's typically a collection of components, including models, tools, prompts, code, and operating procedures. They are combined within the context of a single AI agent for a defined job or role. So an AI agent is somewhat analogous to a human employee or contractor, it has a role, responsibilities, and skills, and uses tools or other resources to accomplish tasks based on the work it is responsible for.

The most important thing to know is that AI agents can collaborate with other AI agents - and human workers. So to accomplish higher-level objectives, AI agents can work together - just like humans do.

Multi-agent systems and AI agent swarms

As a traditional company grows, it's common to see the workforce grow and become more specialized. Also, when headcount increases, the company becomes more complex and requires more management. The same is true when the multi-agent AI systems or swarms grow. The agents generally become more specialized and become increasingly more complex to coordinate. But, AI agents can also manage and organize the work of other AI agents.

So AI agents don't necessarily need to be directed by humans. Instead, one or more AI agents can be managers of other agents (and humans) with knowledge of their skills or capabilities. They can then reason through the best ways to organize and manage other AI agents and humans based on a given objective. The specifics on how this works are beyond the scope of this post but just like with human employees, it mostly depends on the decision-making atonommy the AI agent is provided with.

Different AI orchestration platforms have different ways of implementing multi-agent systems and again, not all of the platforms use the same terminology. But the general concepts are the same. AI agents can be made aware of other AI and human resources, and then they can reason through how and when they need to collaborate. At this point, AI agent discovery and collaboration is mostly a function of the orchestration platform you decide on. However, protocols like Agent Protocol are emerging to make it easier for AI agents to discover and tap the capabilities of other agents. So, it's a pretty safe bet that soon, AI agents will be able to find and collaborate with other AI agents across different platforms.

The main point is that AI orchestration platforms are more than simple workflow automation systems. Their overarching goal is to enable AI agents to reason through complex workflows - without necessarily needing pre-defined (read: human-defied) steps.

Emerging tools and platforms

There are several emerging AI orchestration tools and platforms. The oldest of them is LangChain which was released in October of 2022. Others include AutoGPT, AutoGen, BabyAGI, and CrewAI to name a few. Plus more to come I’m sure.

I've been dabbling with a few different AI orchestration tools. While each provides its unique features and benefits, they all serve a similar purpose: to make it easier to combine multiple AI models, tools, and agents to accomplish things that a single AI model or system can't. I won’t cover specifics today. I'll save that for another post. But, getting familiar with these tools is a must.

A real-world example of AI orchestration

In previous posts, I've discussed AI for writing and publishing books.Book publishing isn't overly complex. It involves several steps, but the steps are generally well understood. So let's look at a few steps to illustrate the general idea behind AI orchestration.

In a traditional case, we might have a single human author who is responsible for both researching and writing a book. Or, the author might work with one or more human researchers. Human subject matter experts could also be involved to ensure the content is accurate and relevant. Then, when a draft is written, a human editor might be used to review and improve the content. There are many more steps, but you get the idea.

In the AI case, we might have one AI agent do the research, another to write the content, and another for the editing. Or we might have a human in the loop for the editing. The AI agents might use a combination of language models, prompts, contextual data, search engines, and other tools or resources to accomplish their respective tasks. Further, each agent might be given explicit instructions or reason through the process itself. In either case, the agent is responsible for a defined purpose, but the specific steps it takes will depend on how the agent is designed and the context in which it operates.

Building multi-agent systems vs traditional business building

The more I think about AI orchestration, the more I see parallels between building multi-agent systems and traditional company building. In both cases, you have defined goals that requires assemblimg a collection of workers to accomplish those goals. In both cases, workers need clear objectives and access to the resources necessary to do their job.

To align with the company mission and vision, the workers need to be able to work together to accomplish the overarching goals. In the same way, AI agents need to be able to work together to accomplish higher-level objectives. And in both cases, the optimal workflow needs to be reasoned through.

You also have to decide which workers to hire and how they should be organized. Weather to define standard operating procedures or let workers reason through processes on their own. And in both cases, you'll continually evaluate and adjust as needed to improve performance.

So, AI orchestration is much more than a technology consideration. And in the same way that assembling the right people, and positioning them in the right roles is essential to building a successful company; assembling the right AI agents and positioning them in the right roles is essential for building a successful multi-agent AI system.

That's it for Now

I think that's enough for today. In a follow-on post, I’ll expand and discuss how blockchain technologies play into things.

Share The Dabble.AI Project

In my last post I said I'd outline the process I'm using for generating books with AI. In this post I'll do that. I've also recorded a video that walks through the process. You will find a link to that in this post as well.

The Problem

When I wrote Exploring GPT-3 it took me 11 months and a little over 1,000 hours to complete. But within a matter of months after it was published, the book was out of date. Of course now, AI can write books in literally a matter of minutes. But when you read those books it's clear they were AI generated. So, I'm trying to find a middle ground. I want to use AI to create the content for a book, but I want to be able to ensure it reads and flows like it was written by an experienced human author. I also don't want to write a single sentence or line of code in the process.

The Solution

What I've come up with is a workflow that essentially lets me read and approve/reject, or ask for revisions on everything that is generated - while the book is being created. So, I don't just let the AI just run through and create an entire book in one shot. Instead, I have it create a list of chapters, then I have it create a list of topics for each chapter, then the content for each topic, and so on.

After each step, I have it get my approval before moving on. This way I can ensure the book flows well, is well written, and accurate while it's being created. The best part is, this process provides incredible results without having to write a single sentence or line of code. So, the book is 100% written by AI, but I'm in full control of the final product.

I walk through the entire workflow in the following video. But I'll also outline the process here in text.

Here's an overview of the high-level steps.

Macro Steps

1. Generate a book overview

2. Generate a list of chapters

3. Save the overview and list of chapters to files named overview.yml and chapters.yaml respectively

4. Make the overview.yml and chapters.yaml file available to the assistant for reference

5. Generate topics for each chapter

6. Generate content for each topic

7. Save the content for each chapter to a file named chapter-#.yaml

8. Make each chapter available to the assistant for reference

9. Have the assistant merge the chapters into a single yaml file named {book-name}-{revision}.yaml

10. Have the assistant use the final yaml file to generate a PDF or some other format for the book

Prerequisites

I've been mostly using OpenAI assistants for this process. But I've also tested it with ChatGPT and Google's Bard and with minor adjustments they both work as well. That said, I prefer using OpenAI assistants. So, to follow this workflow the way I've outlined it, you'll need an OpenAI Platform account. This isn't the same as a ChatGPT account. Again, you could do this using GPTs, but the video walks through how to do it using the OpenAI Platform.

If you don't have a OpenAI Platform account you can sign up here: OpenAI Platform Account.

Creating a new OpenAI assistant

After you've signed up for an OpenAI Platform account, you'll need to create a new assistant. You can do this by selecting the assistants menu item from the left-hand navigation. Then click the "Create Assistant" button.

Assistant Settings

When creating a new assistant, you'll need to provide some basic information about the assistant. I give my assistants a name that is descriptive of the book I'm working on. For example, I named the assistant I'm using for the game theory book, "Game Theory Book Assistant". You can name it whatever you like, it really doesn't matter. What does matter are the settings you choose for the assistant. Especially the instruction.

Instructions

The instructions are where most of the magic happens. They are the starting prompt that provides the assistant with some context and details about what you want it to do. For my workflow, I update the instruction four times. When I first create the assistant, after I've generated the book overview and chapters, after I've generated the content for each chapter, and before I compile the chapters and generate the final output. Below are the instructions I use for each step.

Initial instructions

The first instruction is to generate an overview of the book. After I approve the overview, I'll have the assistant generate a list of chapters. Here are the instructions I'm using for the first step.

You are a book author who is an expert on game theory. You are working on a book named Winning Alone: Game Theory Essentials for Solopreneurs. Your writing style is no-nonsense, direct, and to the point. You write in short sentences and avoid filler words, fluff, and hype. You begin by generating an introduction for the book that is 4000 characters or less. You ask the user if they are happy with the introduction. If they are happy, you write and run code to save the introduction as a file named overview.yaml. If they are not happy, you ask the user for feedback and then generate a new introduction. You continue this process until the user is happy with the introduction.

After the user is happy with the introduction, you generate a list of suggested chapter titles and numbers. You ask the user if they would like to make any modifications. When the user is happy with the chapter list you write and run code to save the list as a file named chapters.yaml.

After I have the overview and chapters, I'll update the assistant's instructions to have it generate topics for each chapter. Before I do that, I'll save the overview and chapters to files named overview.yaml and chapters.yaml respectively. Then I'll make those files available to the assistant for reference. You can do that by uploading the files to the assistant's knowledge base and saving the changes to the assistant.

Instructions to generate the chapter content

After I've saved the overview and chapters to files and made them available to the assistant, I'll update the assistant's instructions to have it generate topics for each chapter. Here are the instructions I'm using for that step.

You are a book author who is an expert on game theory. You are working on a book named Winning Alone: Game Theory Essentials for Solopreneurs. Your writing style is no-nonsense, direct, and to the point. You write in short sentences and avoid filler words, fluff, and hype. You get a list of the book chapters from your knowledge base. Then you ask the user for the chapter they'd like to work on. After the user provide a chapter number you generate a numbered list of topics for the chapter using your knowledge base to get the chapter number and title. You ask the user to confirm the topics are acceptable or to suggest changes. After the user is happy with the topic list you ask the user if they'd like to start generating content for each topic. After generating the content for topic, you make sure the user is happy with the topic content before moving on to the next topic. When you're done with the content for all the topics, you write and run code to save the content for the chapter as a file named chapter-#.yaml.

After each chapter's content is generated and saved to a file, I'll make the file available to the assistant for reference. When I'm done with all the chapters, I'll update the assistant's instructions to have it merge the chapters into a single yaml file named {book-name}-{revision}.yaml.

Instructions to merge the chapters

When I've uploaded all the chapter files and made them available to the assistant, I'll update the assistant's instructions to have it merge the chapters into a single yaml file. Here are the instructions I'm using for that step.

Using the chapters in your knowledge base, you write and run code to merge the chapters into a single yaml file named Winning-Alone-Game-Theory-Essentials-for-Solopreneurs-1.yaml and provide the user with a link to download the file. 

Note: Change the book name to match the name of your book and the revision number to match the current revision number.

After the final yaml file is generated, I'll make it available to the assistant for reference. Then I'll update the assistant's instructions to have it generate the final output.

Instructions to generate the final output

Here are the instructions I'm using to have the assistant generate the final output.

Using the Winning-Alone-Game-Theory-Essentials-for-Solopreneurs-1.yaml file in your knowledge base, you write and run code to generate a PDF of the book and provide the user with a link to download the file. 

Assistant Settings

Aside from the instructions, there are a few other settings you'll need to configure for the assistant. Here are the settings I'm using.

Model

• Use the most recent model. At the time of this writing, that's the gpt-4-1106-preview model.

Tools

• Enable the "Code interpreter" tool

• Enable the "Retrieval" tool

Files

You will add the files as you generate them. The files you'll create are:

• overview.yaml

• chapters.yaml

• chapter-#.yaml (where # is the chapter number)

• {book-name}-{revision}.yaml

After adding each file, be sure to save the changes to the assistant.

Conclusion

That's it. That's the process I'm using to generate books with AI. It's a bit more involved than just letting the AI run through and generate an entire book in one shot. But I think the results are worth it. I'm able to ensure the book reads and flows well, is well written, and accurate while it's being created. And I don't have to write a single sentence or line of code in the process.

I hope this process is helpful. If you have any questions, feel free to ask in the comments below.

My goal for this project is to figure out how to leverage AI to plan, build, launch, and grow a profitable company that has only one employee - me. Even though my company Dabble Lab has incredibly talented people that could help me out, I want to do everything myself. I'm doing this to hopefully prove it's possible.

I'm convinced that in the not so distant future, anyone will be able to easily use AI to plan, build, and launch companies that are also mostly run by AI. The human role will be primarily focused on vision, strategy, and oversight. In recent history, the norm has been to work with, or for someone else. That's changing. It's been changing for some time. But AI is going to impact that change in a big way. Like, I fully expect that we'll see billion dollar companies with only one employee in the next 5-10 years. Maybe sooner.

The general trajectory is that AI will continue to improve, enabling humans to do more and more, until almost everything can be done by a single person. I'm not saying that's going to happen overnight. Or that it will happen for every type of business. But it's going to happen. It's just a question of how long it takes.

Who will create the first billion-dollar companies with no employees?

I don't know. But my guess is that it won't be the type of person we might expect. The idea of using AI to create a billion-dollar startup seems like something that is going to take a lot of technical expertise. Maybe not. From my perspective, a deep understanding of technology isn't all that important. What's important is understanding the problem you're trying to solve. And then being creative enough to figure out how to use AI to solve it.

I'm a technical person. I've been writing code and building software for over 30 years and working with AI for over 10 years. But I'm realizing that my past experience isn't all that helpful for this project. In fact, at times, it's a hindrance. Because the ways I would normally approach a problem are not always the best approach when AI is involved. I catch myself making three left turns to go right. Said another way, I find myself over complicating things. I'm also finding that too often I underestimate what AI can do. Here's an example.

As I've shared in previous posts, I'm working through how I can use AI to write a book about Game Theory. It's not as simple as just saying to ChatGPT - "Write me a book about game theory". I've tried that. It doesn't work. But if I break the process down into smaller steps, it does work. For example, I can ask ChatGPT to write an outline for the book. Then an outline for each chapter with bullet points for topics that should be covered. Then I can have ChatGPT write content for each bullet point. 

This is a lot of manual prompting. But it works. And it works well. So, of course, as a software guy, my mind goes to writing code to automate the process of submitting all the prompts. But I'm trying to be mindful about my goals for this project - to have AI to do everything. So, I'm resisting the temptation to write code. I could just suck it up and submit all the prompts manually. But again, my goal is to have AI do as much as possible. Plus, that's way more time than I want to spend. So, I'm forcing myself to ask how I can get AI to do it - anytime I’m faced with something time consuming or that requires a lot of human expertise. 

Of course, my first thought was to just have AI write the code to automate the process of submitting all the prompts. That’s pretty easy to do. But then I started wondering if there was a way to get ChatGPT to prompt itself. I've dabbled with projects like AutoGPT and BabyAGI that are using this general approach. But those tools are code dependent and require a reasonable amount of technical skills to get them up and running. What I wanted was just a simple way to get ChatGPT to walk through a prompting process for generating a book. Without any code or external tooling.

I got it working. And it works surprisingly well. I'll share details in a future post. But the point is, I'm having to force myself to think differently. To almost unlearn years of training and conditioning. It's not been easy for me. But I'm sure it will feel natural for some. That's why I'm speculating that the first billion-dollar companies with no employees could likely be created by people that don't have a technical background. Maybe people without a lot of business experience either.

That’s it for now. Thanks for reading and stay tuned. The plan for my next post is to outline the book generation process.

Share The Dabble.AI Project

There are at least two paradoxes I see for AI business opportunities.

1. The best opportunities should not be the focus.

2. The biggest opportunities should not be the focus.

I’m defining the best opportunities as the ones that are clearly a perfect fit for AI. Just search something like “make money with AI” on YouTube and you’ll see a bunch of them. Generally speaking, if it’s easy and inexpensive to do with AI, everyone will copy the idea as soon as the opportunity presents itself. So, unless you're the first mover, or a very early mover - these won’t be opportunities long. So they are not worth focusing on - IMO.

The biggest opportunities are the ones with the biggest market caps and/or the most strategic value (on a global level). Like developing AI tools for writing code. Or of course, developing AGI. These are massive opportunities for sure. But they require massive investments. So they are not likely to be viable opportunities for solopreneurs (the focus of The Dabble.AI Project). Meaning, they aren’t worth focusing on either.

So, what is worth focusing on?

Here’s what I’m looking for personally. Opportunities that have the following attributes.

1. The opportunity can be tested quickly

2. Personal knowledge / experience is required

3. Its not obvious to others and not easily spotted if successful

4. The TAM (total addressable market) is relatively small

5. Barriers to entry for future competitors could be created

Let me break my thinking.

1. The opportunity can be tested quickly

Speed to market should be a top priority. Product life cycles are getting shorter and shorter. So, spending years developing something before being able to validate the potential is a bad ideas - IMO. In my case, if I can’t have something to market in 180-days or less - I’m not interested.

2. Personal knowledge / experience is required

Humans and machines both have strengths and weaknesses. Either working on a task that is best suited to their strengths will outperform the other. But, when they collaborate, they can outperform either working alone. So, I’m most interested in business models that require a combination of AI and human intelligence - but not human time. This mostly boils down to business models that I feel qualified to oversee.

3. Its not obvious to others and not easily spotted if successful

With artificial intelligence, business models will be so easy to execute on that competition will exist as soon as an opportunity is known. For that reason, a competitive advantage could be gained by keeping business operations as covert as possible. So, telling the World how well you’re business is doing on social media might not be the best plan.

4. The TAM (total addressable market) is relatively small

Small markets tend to draw less attention from bigger competitors. You’re also more likely to be able to capture a large percentage of a smaller market faster. It could also make sense to have small slices of larger markets. But given the option - I’d prefer a large slice of a smaller market.

5. Barriers to entry for future competitors could be created

This could be the single most important consideration. If or when someone else decides to use AI to compete - what’s your competitive advantage? What things can you do to make it harder for them? These can be things like using custom models that are trained with data that isn’t easily accessible, or capturing a large share of the market.

I’m also making the assumption that there aren’t any perfect opportunities. So, the five points above aren’t “in stone”. They are just a wish list of things like love to see in the business models I’m looking for. Of all those items, #1 and #5 are the most important to me.

Today I spent about 5 hours working on a second draft of the game theory book. I didn’t spend any time writing or editing - the book is 100% AI generated at this point. All my time went into tweaking prompts and working through the prompt chaining.

This draft is still not something I’d publish. But it’s moving in the right direction - quickly. It’s 100x better than the first draft - in my opinion. I’m trying to get it to sound less like a Wikipedia page - it’s getting there. But to get it sounding more natural - it’s going to need more prompt tweaks. To that end, I had an idea - and I need your help.

So here’s my idea. I’m thinking that I could collect feeback using the comments on Google Docs, then feed the comments into a ChatGPT prompt that would be used to to make revisions for the next draft.

So, I have a favor to ask.

Would you mind reviewing the second draft and leaving comments about anything that you think needs addressing? You can skip to any chapter you’d like and provide as much or as little feedback as your time permits.

I’m not sure this idea will work but I’m super curious to find out.

If you do provide feedback…

• I’ll list your name as a reviewer in the final version of the book - if you’d like.

• I’ll also give you access to the prompts I used to generated the book - if you want.

Thanks in advance for your help!

Review Draft 2

Today is December 19, 2023. It's 7:07 AM. It's a Tuesday. I've been working on The Dabble.AI Project for 10 days now. I’ve spent just over 20 hours and all I have are a bunch of notes and a few blog posts to show for it. But my main objective isn’t to create a blog. I'm trying to create an automated business. There is a fine line between planning and procrastinating.

It took me almost 12 months to write my book Exploring GPT-3. I invested just over 1,000 hours total. I think I could write the same book today in a matter of weeks - in under 100 hours - using ChatGPT and a few other tools. I might be able to do it in a few days if I didn’t need to spend time on research, testing, editing, and formatting. My Exploring GPT-3 book has sold very well. But, I invested a lot of time and the royalties percentage I get from the publisher isn’t great. So, it has not been a super exciting ROI from a time invested perspective. However, if I was able to create a book in 100 hours or less, and publish it myself - maybe that would make it more interesting. 

Assuming I’m charging myself $125/hour, if I could use AI to create and publish a book in under 100 hours, my time investment would be $12,500. If the book sold 625 copies at an average of $20 / copy - I’d break even. But that assumes I don’t spend time or money on anything else - like marketing / promoting the book. That’s not realistic. Let’s plan $5,000 for marketing and misc. So, total investment would be $17,500. Meaning, I’d need to sell 875 copies to break even. I should test that theory. But spending 100+ hours is more than I really want to invest on this model. At this point anyhow. Maybe I should try to create a book in a few hours just to dabble through the workflow. I’ll do that. I’m going to spend one hour trying to create a book. I’ll decide from there if I want to spend more time.

One of my favorite topics is game theory. I’ve been thinking about it a lot lately. I was going to write a post about it. But again, I’m not trying to create a blog. So, maybe I’ll try to use AI to generate a book about game theory. I’m guessing if I search Amazon there are tons of game theory books already. So, lots of competition. I’m also guessing the target market for people just interested in learning about game theory isn’t big - relatively speaking. I could use AI to try and gauge the market size but I’m more interested in seeing how fast I could get a book to market first. If the product is decent, and I can get it done quickly - I’ll invest more time on market research. 

How should I engineer the prompt? I need a title, subtitle, short description, outline, chapter titles, chapter topics, and the chapter content. From there I need a list of other stuff I’ll need to get done. Like creating a cover image, getting an ISBN number, and formatting it for Amazon. 

I’ve never self published a book before. Packt Publishing did all of this for my Exploring GPT-3 book. So, that part is a black box for me. I see tons of self-published books so I assume the process is simple. Not sure if it can be automated but I would bet it probably can. But that’s not important now.

Let’s start with a prompt. I’ll use GPT 4 because it’s pretty good with more complex instructions. I want GPT to break things down into steps/parts so I can provide feedback as it generates the book. I’m not going to spend time trying to optimize the prompt now because I want to see how far I can get in an hour. So, here’s the prompt I’ll use.

I want to try an create a book in an hour. I want to write it on a topic I know well so that I can read and review it quickly. The topic will be game theory. I want to start by coming up with a title, subtitle, and description. From there I'd like to create an outline. The outline should include a chapter list and a few bullet points for each chapter. After confirming that I like the outline, write each chapter but confirm that I like the chapter before moving on to the next. Once the book is written, list the other things I'll need to publish the book on Amazon. For example, I'll need a cover image and I'll need to get the book into a format that works for Amazon. Let's get started. 

Here are the results (aka: completion) from my first prompt.

Fantastic! Embarking on a rapid book-writing journey on game theory sounds like an exciting and intellectually stimulating endeavor. Here's a structured approach to kickstart your project:

### Step 1: Title, Subtitle, and Description

1. **Title**: "Strategic Moves: Mastering Game Theory"
2. **Subtitle**: "A Concise Guide to the Art of Rational Decision Making"
3. **Description**: This book offers a swift yet comprehensive dive into the world of game theory. Designed for enthusiasts, students, and professionals alike, "Strategic Moves" demystifies the complex principles of game theory and applies them to real-world scenarios. From classic dilemmas like the Prisoner's Dilemma to advanced concepts in evolutionary game theory, this guide serves as a practical companion for anyone looking to enhance their strategic thinking skills.

### Step 2: Outline with Chapter List and Bullet Points

1. **Introduction to Game Theory**
   - Definition and historical background
   - Basic concepts: players, strategies, and payoffs
   - The importance of game theory in decision making

2. **Key Concepts and Terminologies**
   - Nash Equilibrium
   - Dominant strategies and Pareto efficiency
   - Zero-sum games vs non-zero-sum games

3. **Classic Games and Dilemmas**
   - The Prisoner's Dilemma
   - The Stag Hunt
   - The Battle of the Sexes

4. **Game Theory in Economics and Business**
   - Auctions and bidding strategies
   - Market competition and oligopolies
   - Contract theory and mechanism design

5. **Game Theory in Politics and International Relations**
   - Electoral strategies and voting systems
   - Deterrence and arms races
   - Coalition building and negotiation

6. **Evolutionary Game Theory**
   - Evolution of cooperation
   - Social norms and conventions
   - Game theory in biology and ecology

7. **Behavioral Game Theory**
   - Irrational behaviors and bounded rationality
   - Trust, reputation, and fairness
   - Experimental game theory

8. **Applications and Case Studies**
   - Applying game theory in everyday life
   - Historical case studies
   - Future directions and new frontiers

9. **Conclusion**
   - Summary of key learnings
   - How to continue exploring game theory
   - Final thoughts and reflections

### Step 3: Confirmation and Writing

Let's confirm if you're happy with this outline. Once confirmed, we'll proceed to write each chapter, ensuring you're satisfied before moving to the next.

### Step 4: Preparing for Publication

After writing, we'll need to:
- Create a cover image
- Format the manuscript for Amazon (e.g., Kindle Direct Publishing guidelines)
- Consider additional elements like ISBN, author bio, and a compelling book blurb

Do you want to proceed with this outline, or are there any adjustments you'd like to make?

GPT-4 did a good job following all the instructions I provided in my prompt. I don’t love the title but I like the sub title. I’ll go back and work on a better title later. The description is pretty good. There aren’t enough chapter topics so I suspect the chapters will be too short. That’s because of the sentence: “The outline should include a chapter list and a few bullet points for each chapter.” I should not have limited the bullet points to a few. But I’m not spending more than an hour so I’ll deal with that later if I decide to move forward. So for now, I'm going with it as is.

My next prompt is:

Yes, let's proceed with this outline

ChatGPT gets to work on the first chapter and comes back with.

Great! Let's begin writing the first chapter. Since each chapter is a substantial piece of writing, I'll provide a concise version that you can expand upon.

---

# Chapter 1: Introduction to Game Theory

**Definition and Historical Background**
Game theory, a field of applied mathematics, provides a framework for analyzing situations in which parties, known as players, make decisions that are interdependent. This interdependence causes each player to consider the other's possible decisions or strategies when making their own. The origins of game theory can be traced back to the early 20th century, with significant contributions from John von Neumann, John Nash, and others.

**Basic Concepts: Players, Strategies, and Payoffs**
In game theory, a 'player' can be an individual, a group of individuals, or even a biological species. A 'strategy' is a complete plan of action a player will follow in a given situation. A 'payoff' is the outcome of a strategy, often quantified in terms of utility or some form of benefit or cost.

**The Importance of Game Theory in Decision Making**
Game theory has become an essential tool in economics, political science, psychology, and biology. It helps in understanding and analyzing strategic interactions in competitive and cooperative environments. From understanding economic markets and political elections to studying evolutionary biology, game theory provides insights into how and why individuals and groups make decisions.

---

Let me know if you're satisfied with this brief version of Chapter 1, or if you have any specific changes or additions in mind. Once approved, we'll proceed to the next chapter!

As I suspected, the chapter is too short, but not a bad start. I’m just going to continue.

Yes, let's proceed with this chapter

Here is what I get back for chapter 2.

Great! Let's move on to Chapter 2.

---

# Chapter 2: Key Concepts and Terminologies

**Nash Equilibrium**
The Nash Equilibrium, named after mathematician John Nash, is a fundamental concept in game theory. It represents a situation where no player can benefit by changing their strategy while the other players keep their strategies unchanged. This concept is crucial for understanding the stability of the strategies chosen by the players in various games.

**Dominant Strategies and Pareto Efficiency**
A dominant strategy is an action that is the best choice for a player, no matter what the other players do. On the other hand, Pareto efficiency, named after economist Vilfredo Pareto, occurs when no player can be made better off without making at least one player worse off. This concept is used to analyze the efficiency and desirability of different outcomes in a game.

**Zero-Sum Games vs Non-Zero-Sum Games**
In zero-sum games, one player’s gain is equivalent to another player’s loss, so the total gain or loss in the game is zero. Examples include games like chess or poker. In contrast, non-zero-sum games are scenarios where the sum of gains and losses can be more than zero, which is typical in most real-world situations. These games illustrate the potential for cooperative strategies where all players can benefit.

---

This chapter introduces the reader to some of the core concepts and terminologies in game theory. Let me know if this aligns with your vision for Chapter 2, and if you're ready to proceed to the next chapter or if there are any adjustments you'd like to make.

It's now 7:34 AM - I've spent 27 minutes on this so far. I’ve been trying to copy everything into VS Code to formate everything in markdown for a blog post. That slowing me down. I’ll just move on in ChatGPT and worry about getting a blog post done later - or maybe not.

Moving on.

Yes, let's proceed with this chapter

The completion for chapter 3 was:

Fantastic! Moving on to Chapter 3.

---

# Chapter 3: Classic Games and Dilemmas

**The Prisoner's Dilemma**
The Prisoner's Dilemma is a standard example of a game analyzed in game theory that shows why two completely rational individuals might not cooperate, even if it appears that it is in their best interest to do so. It is a fundamental problem in understanding trust, cooperation, and decision-making between parties.

**The Stag Hunt**
The Stag Hunt is another important game in game theory that demonstrates how cooperation and trust play a role in achieving the best outcome. It contrasts with the Prisoner's Dilemma by showing a situation where mutual cooperation is the best strategy for all, but this requires trust that the other party will not defect.

**The Battle of the Sexes**
This game illustrates a conflict between two parties, each preferring different outcomes. It is used to demonstrate coordination games, where players must coordinate their strategies for a more favorable outcome but have different preferences on which outcome to coordinate on.

---

These classic games are essential for understanding the various dynamics at play in game theory. They provide foundational knowledge for grasping more complex scenarios and applications. Let me know if this chapter meets your expectations and whether you're ready to move on to the next one or if there are any changes you'd like to see.

I'm going to continue with the next chapter.

Yes, let's proceed with this chapter

Here is what I get for chapter 4:

Excellent! Now we'll delve into Chapter 4, focusing on game theory in the context of economics and business.

---

# Chapter 4: Game Theory in Economics and Business

**Auctions and Bidding Strategies**
In the world of economics and business, auctions play a significant role. Game theory helps in understanding various auction formats and bidding strategies. It explores how bidders strategize to win auctions while trying to pay as little as possible, from sealed-bid auctions to English and Dutch auctions.

**Market Competition and Oligopolies**
Game theory is instrumental in analyzing market competition, especially in oligopolistic markets where a few firms dominate. It examines how companies strategize regarding pricing, product differentiation, and market expansion, considering the potential reactions of competitors.

**Contract Theory and Mechanism Design**
Contract theory, a part of game theory, explores how economic actors construct contractual arrangements, generally in the presence of asymmetric information. It is vital in designing mechanisms or systems that lead to desirable outcomes by aligning the interests of different parties, such as in auctions, voting systems, or incentive structures.

---

This chapter provides a practical application of game theory in real-world economic and business scenarios, demonstrating its value in strategic decision-making and policy design. If you're satisfied with the content and structure of this chapter, we can proceed to the next one. Please let me know if there are any modifications or additions you would like.

I really didn’t need to have it prompt me for a confirmation after each chapter since I’m just rolling through the process. But if I was spending more time. This would be the right way to go.

Anyhow, moving on.

Let's continue

The completion was:

Great, moving on to Chapter 5, which explores the application of game theory in politics and international relations.

---

# Chapter 5: Game Theory in Politics and International Relations

**Electoral Strategies and Voting Systems**
Game theory offers valuable insights into electoral strategies, helping to understand how candidates and parties choose their platforms and campaign strategies. It also sheds light on the effectiveness and implications of different voting systems, such as first-past-the-post and proportional representation, on election outcomes.

**Deterrence and Arms Races**
In international relations, game theory is crucial for understanding deterrence and arms races. It helps analyze how nations develop strategies around the development and deployment of military resources, balancing the desire for power with the risks of escalation and conflict.

**Coalition Building and Negotiation**
Game theory also applies to the formation of coalitions in politics and the dynamics of international negotiations. It explores how parties with different interests can find common ground, the dynamics of cooperation and conflict, and the strategies for effective negotiation and coalition-building.

---

This chapter provides insights into the strategic complexities in the realms of politics and international relations through the lens of game theory. If this chapter aligns with your expectations, we can move forward to the next one. Please let me know if you'd like any changes or additional content.

I'm going to continue with the next chapter.

Yes, let's proceed with this chapter

Here is what I get for chapter 6:

Great! Let's continue with Chapter 6, focusing on evolutionary game theory.

---

# Chapter 6: Evolutionary Game Theory

**Evolution of Cooperation**
Evolutionary game theory extends traditional game theory into the study of evolution and natural selection, particularly focusing on how cooperative behavior can evolve in populations. It examines the conditions under which cooperation can emerge and be sustained among self-interested individuals, a question crucial to understanding social behaviors.

**Social Norms and Conventions**
This aspect of game theory explores how certain behaviors become social norms and conventions. It looks at how these norms evolve, are enforced, and sometimes change over time, providing a framework for understanding the development of societal rules and the role of individual actions in shaping social dynamics.

**Game Theory in Biology and Ecology**
In biology and ecology, game theory models are used to understand and predict the behavior of various species in their struggle for survival. This includes analysis of predator-prey dynamics, mating strategies, and resource competition, offering insights into the adaptive behaviors of different organisms.

---

Evolutionary game theory provides a fascinating perspective on the natural world and human society, bridging the gap between biological processes and social interactions. If you're satisfied with this chapter and ready to move forward, please let me know. Alternatively, if there are specific aspects you'd like to add or modify, feel free to mention them.

Moving on.

Yes, let's proceed with this chapter

And, here is what I get for chapter 7:

Excellent! Now, we'll proceed to Chapter 7, which delves into behavioral game theory.

---

# Chapter 7: Behavioral Game Theory

**Irrational Behaviors and Bounded Rationality**
Behavioral game theory challenges the assumption of complete rationality in traditional game theory. It examines how cognitive biases, emotions, and limited cognitive resources (bounded rationality) affect decision-making. This perspective is crucial for understanding real-world decisions, which often deviate from what is predicted by classical rational models.

**Trust, Reputation, and Fairness**
Trust and reputation play significant roles in many strategic interactions. Behavioral game theory explores how these factors influence decisions, particularly in repeated games. It also delves into concepts of fairness and how individuals' perceptions of fairness affect their strategic choices, even sometimes leading them to act against their own self-interest for the sake of equitable outcomes.

**Experimental Game Theory**
Experimental game theory involves conducting experiments, often with human subjects, to study strategic interaction in controlled settings. These experiments help in testing and refining theoretical predictions of game theory, providing empirical evidence about how people actually behave in strategic situations.

---

This chapter brings a human element to game theory, considering how real people behave in strategic situations, which can be quite different from theoretical predictions. Let me know if this chapter captures the essence of what you envisioned for behavioral game theory. If you're ready to proceed to the next chapter or if there are any adjustments or additions, please share your thoughts.

Still just continuing with the next chapter.

Yes, let's proceed with this chapter

Here is the completion for chapter 8:

Great, let's move on to Chapter 8, which will focus on the applications and case studies of game theory.

---

# Chapter 8: Applications and Case Studies

**Applying Game Theory in Everyday Life**
Game theory isn't just an academic discipline; its principles can be applied in everyday decision-making. This section explores various real-life scenarios, from negotiating salaries to making purchasing decisions, where game theory principles can be applied for more strategic and effective outcomes.

**Historical Case Studies**
To illustrate the practicality of game theory, this chapter includes a selection of historical case studies. These might include critical moments in economic history, political decisions, and even famous military strategies, analyzed through the lens of game theory to provide a deeper understanding of the strategic choices made.

**Future Directions and New Frontiers**
The field of game theory is continually evolving, and this section explores emerging trends and potential future applications. This includes its growing influence in areas like technology, cybersecurity, artificial intelligence, and complex system analysis, highlighting how game theory is likely to shape decision-making in various domains in the future.

---

This chapter aims to bridge the gap between theory and real-world application, showcasing the versatility and relevance of game theory. If you're satisfied with the content and direction of this chapter, we can wrap up with the conclusion next. Please let me know if there are any changes or additional topics you'd like included.

And now for the last chapter.

Yes, let's proceed

Here is the final chapter:

Fantastic! We'll now conclude with Chapter 9.

---

# Chapter 9: Conclusion

**Summary of Key Learnings**
This concluding chapter will recapitulate the main concepts covered throughout the book. It will summarize the key principles of game theory, its applications in various fields, and the insights it provides into strategic decision-making, cooperation, competition, and human behavior.

**How to Continue Exploring Game Theory**
For readers interested in delving deeper into the subject, this section will provide guidance on further resources and study paths. It will include recommendations for advanced books, online courses, and other educational materials to continue exploring the vast and fascinating world of game theory.

**Final Thoughts and Reflections**
The book will close with reflections on the importance of understanding game theory in today's interconnected and complex world. It will emphasize how the principles of game theory can be applied in personal, professional, and societal contexts to make more informed and strategic decisions.

---

With this conclusion, your book on game theory would be complete in its content. Let me know if this conclusion aligns with your vision, or if there are any final thoughts or messages you'd like to include.

After this, the next steps would involve preparing your manuscript for publication, including formatting, cover design, and other publishing logistics.

Now, because of my initial prompt, I get a few suggestions for next steps. So, I'll ask what I need to do to prepare it for publishing on Amazon.

What do I need to do to prepare this book for publishing on Amazon?

The completion was:

Publishing a book on Amazon, particularly through Kindle Direct Publishing (KDP), is a straightforward process but requires attention to several key details. Here’s a step-by-step guide to get your book ready for Amazon:

### 1. Finalize Your Manuscript
- **Editing and Proofreading**: Ensure your book is thoroughly edited and proofread. This includes checking for grammatical errors, consistency in style and formatting, and overall flow of the content.

### 2. Format Your Manuscript
- **KDP Formatting**: Amazon has specific formatting requirements. Your manuscript should be formatted according to KDP’s guidelines, which include margins, font, line spacing, and chapter headings. KDP provides templates to make this easier.
- **File Format**: The preferred format for Amazon is a Word document (.doc or .docx), but you can also submit your manuscript in other formats like PDF.

### 3. Design a Book Cover
- **Professional Cover Design**: The book cover is crucial as it's the first thing potential readers see. You can design it yourself or hire a professional. The cover must meet Amazon’s dimension and quality requirements.
- **Cover Art**: If you’re using images, ensure you have the rights to use them. KDP offers a Cover Creator tool with pre-made designs and royalty-free images.

### 4. Create a Compelling Book Description
- **Engaging Description**: Write a description that clearly summarizes your book and entices readers. Include key points and what readers can expect to learn.

### 5. Consider Additional Elements
- **ISBN**: While Amazon can provide a free ISBN for your book, getting your own ISBN gives you more control and wider distribution options outside of Amazon.
- **Author Bio**: Prepare a brief author biography that will appear on your book’s Amazon page.

### 6. Upload to Kindle Direct Publishing
- **Create a KDP Account**: If you don’t already have one, create a KDP account.
- **Upload Manuscript and Cover**: Follow the steps on KDP to upload your manuscript and cover.
- **Set Your Book Details**: This includes the title, description, keywords, categories, and author information. Good keywords and categories help your book become more discoverable on Amazon.

### 7. Choose Distribution Channels and Pricing
- **Select Distribution Channels**: Decide whether to distribute your book globally or in specific territories.
- **Set Pricing**: Determine the price of your ebook and paperback. Consider your book’s length, genre, and target audience. KDP also offers a royalty calculator to estimate earnings.

### 8. Publish and Promote
- **Review and Publish**: Once everything is set, review your book details and hit publish. It typically takes 24 to 72 hours for the book to appear on Amazon.
- **Marketing and Promotion**: Plan for marketing your book. This can include social media promotion, email marketing, getting reviews, and possibly creating a website or blog.

### 9. Monitor and Update
- **Sales Dashboard**: KDP provides a dashboard to track sales and royalties.
- **Update as Needed**: You can make changes to your book and republish if needed.

Remember, self-publishing requires not only writing but also marketing and promoting your book to reach a wider audience. Good luck with your publishing journey!

Again, I've never self published a book on Amazon before. But it looks like I need to get everything into a Word doc. I'll do that manually by copying and pasting everything into a Google Doc, and then I’ll export the Google Doc to a Word document.

For this to be fully automated, I'd need a way to merge all the chapters into a single document using APIs or something. I'm sure there is a way to do that. I'll look into it later.

It's now 7:57 AM. I'm going to copy and paste the chapters into Google Docs, export it Word, and then call it a day.

I didn't get the book done in an hour but it's a good start. I'll come back to it later.

Here’s the book after 1-hour. It’s not great but it’s a starting point.

8:06 AM - stopping.

Can AI autonomously build and run a profitable business without employees? That’s the question the Dabble.AI project is trying to answer.

In this post, I’ll be outlining the steps I've been testing to generate automatable business models using LLMs. I also explain my working assumptions and the general thinking behind each step.

My wife sometimes tells me "Your mind is like a drunk baby." It's a funny line from the movie Office Christmas Party - so she's not being mean. She's just reminding me that sometimes the things I say can come out sounding random and disconnected. I know she's right - I often confuse myself. I'm trying to stay mindful of that in these posts.

To that end, I’m going to begin each post with a few questions I’m focusing on. But I won’t necessarily have answers. They are more about staying focused. Hopefully that will keep me from sounding completely incoherent. But no promises. 😵‍💫

Focusing questions

Can LLMs be used to discover automatable business models? Can they help me discover models that are unique and not totally obvious to everyone? How? What are the steps? What are the prompts? What LLMs, or combination of tools would do the best job?

My working assumptions

AI tools currently exist to enhance and optimize almost every facet of a company. The tools will get better and better until humans are only needed for oversight. Today, there is a human in the loop for most tasks. That won’t be necessary for much longer. I’d guess by 2033 for online only businesses. Possibly much sooner. At that point, AI will do all of the following things - autonomously.

1. Identify market opportunities

2. Create and test products based on market opportunities

3. Raise funding if needed

4. Create product marketing materials

5. Build and manage online communities

6. Test and execute marketing campaigns

7. Take and fulfill orders

8. Provide product support

9. Monitor and adapt to market demands

10. Iterate and optimize everything

But many offline businesses will also be fully automated. Even in cases where human capital is required, AI will be able to fully manage the workforce. For example, consider a future in-home nursing care business. AI will:

1. Post job openings

2. Interview applicants

3. Conduct background checks on applicants

4. Negotiate compensation based on labor supply and demand

5. Monitor that workers are showing up (GPS tracking - like what Uber does)

6. Monitor customer satisfaction. Think Amazon Alexa devices in the home that does advanced sentiment analysis on the interactions between the caregiver and the person receiving the care.

7. Adjust compensation or replace workers based on customer satisfaction scores

That was an in-home nursing service example but you can apply the same thinking to almost any other in-home services - plummers, electricians, painters, HVAC services - almost anything.

The technology to do everything above already exists. Now it’s just a matter of connecting the dots. That’s often easier said than done. Things like privacy concerns and other compliance nuances will create friction for sure. So, it’s going to take some time. But, if the current technological trajectory continues, this is just a matter of time.

I’m sharing these assumptions because they influence how I’m going about my business model generation research. I’ll expand in future posts. For now however, let’s get into the steps I’m testing and the tools I’m using.

Business model generation steps

Here are the general steps I follow to generate business models with LLMs.

1. Decide on a general business model.

2. Create a list of my personal values, experiences, interests, and skills.

3. Define very specific goals and objectives.

4. List your available resources.

5. Create a prompt using the stuff from steps 1-4.

6. Use the prompt to generate a list of business ideas using multiple LLMs.

7. Identify similar models from each LLM and combine them into a single business model.

8. Prompt each LLM to expand on the combined business model.

Now let me break each step down in more detail.

Step 1: Decide on a general business model.

In my last post, I provided a list of general online business models. The fist step in the process is to choose one of those models. I'll use that model as the basis for the business model I'm going to generate.

NOTE: I’ve tested prompts without doing this. It’s helpful for brainstorming but tends to make the results too broad.

Step 2: Create a list of personal values, experiences, interests, and skills.

In a future post, I'll break down my thoughts here in more detail. But the purpose for this step is two-fold. First, by including your personal values, experiences, interests, and skills, you're more likely to generate business models that you're passionate and knowledgeable about. Second, you're more likely to generate business models that might not be obvious to others. A good business model is NOT one that everyone else is already doing. Or one that anyone else can easily copy. As AI becomes more capable and mainstream, it will be increasingly easier for business models to be copied. So, you want to focus on trying to identify models that are unique, not obvious, and not easily replicated. Again, I'll go into more detail in a future post.

Step 3: Define very specific goals and objectives.

I define goals as outcomes I'd like to see happen, but don't have 100% control over. For example, I might set a sales goal of $1 million dollars. But I don't have 100% control over whether or not I’ll generate $1 million dollars because I can't control the actions, preferences, or decisions of others.

Objectives on the other hand are things I have 100% control over. For example, I can set an objective to write 10 blog posts. In the context of an automated business, I might want the business to generate 1M in revenue but I can't automate that outcome. I can however automate the process of writing 10 blog posts.

The main thing for this step is to be very specific about what you’d like to achieve (goals) and what need to get done (objectives).

Step 4: List your available resources.

This step is pretty straight forward. I create a list of the resources I have available to me. This includes things like capital to invest, time, skills, and relationships, etc. I’m very specific here. I don’t want to generate a business model that requires $500,000 in startup capital if I only have $10,000 to invest. Or for a market I know nothing about and don’t have any contacts in.

Step 5: Create a prompt using the stuff from steps 1-4.

I use the information from steps 1, 2, 3, and 4 to create a prompt. I’m not sharing the prompts I’m testing now for two reasons. First because I’m still testing and refining this process. Second, because I’m using information that others have provided me for testing purposes. So, it’s the secret sauce. Reread step 2 if you’re still wondering why I’m not sharing those prompts. 🤔

NOTE: I can create some example prompts with fake data if that’s of interest to enough people. If so, please leave a comment.

Step 6: Use the prompt to generate business models ideas using multiple LLMs.

After creating the prompt I test it using three or more different LLMs. My go-tos have been ChatGPT, Bard, and Claude.ai. But I’m also testing with:

• AI21 Labs - Jurassic-2 Ultra

• Cohere - Command

• Meta - Llama 2 Chat 70B

Note: If there is a model you think I should be testing with. PLEASE leave a comment.

Step 7: Identify similarities and combine them into a single business model.

Using the results generated by each of the LLMs, create a single business that combines similar results. At this point you should have just one business model.

Step 8: Prompt each LLM to expand on the combined business model.

Using the final business model from Step 7, create a new prompt to expand on the model. For example, that prompt might start something like.

I'm working on a business model for an online business. Can you expand on what I have so far? My current model is:

[model from step 7 goes here]

What's next?

I’ve been thinking a lot about game theory. Questions like: How  would I define the game? How do you win? What are the rules? Who are the players? What are the constraints, paradoxes, risks, rewards, incentives, penalties, and strategies? Mostly about strategies. I’ll probably talk about that next - if I can do so coherently.

Thanks for following.

My objectives are:

1. Use AI tools to plan, build, launch, and run a profitable business.

2. Spend as little human time as possible to achieve my first objective.

3. Document the entire process and share the it with the world.

The Macro Strategy

Here are the high-level steps I have defined.

1. Choose a business model

2. Decide what to sell

3. Define the target audience

4. Develop a go-to-market strategy

5. Create the product

6. Launch the product

7. Test to determine product-market fit

8. Automate or eliminate manual processes

9. Scale and maintain profitability

10. Repeat the process with another product

Step 1: Choose a business model

The first step will be to choose a general business model. The focus of this project is on starting online businesses. So I'll start by choosing one of the following general online business models.

1. E-commerce: Selling goods online through individual websites or platforms like Amazon, eBay, etc. This model includes both B2C (business-to-consumer) and B2B (business-to-business) approaches.

2. Subscription Services: This model can include any service that charges a recurring fee, like SaaS services (Software as a Service), online magazines, or even subscription-based food delivery services.

3. Freemium Model: Offering basic services for free while charging for premium features. This model is common in software services, mobile apps, and online games.

4. Affiliate Marketing: Earning commissions by promoting other people's (or company's) products. This includes bloggers, YouTubers, and social media influencers who link to third-party products.

5. Online Marketplaces: Platforms where third-party sellers can list and sell products, like Etsy for handmade goods, or Airbnb for accommodation rentals.

6. Online Education and Courses: Offering educational content, courses, or training programs online. This can range from platforms like Coursera or Udemy to individual experts selling courses on their websites.

7. Content Monetization: This includes bloggers, podcasters, and video creators generating revenue through ads, sponsorships, memberships, or donations (e.g., using platforms like Patreon).

8. Digital Advertising: Selling advertising space on websites or social media platforms. This includes both direct ad sales and programmatic advertising.

9. Dropshipping: Selling products without holding inventory, where the retailer transfers customer orders to a third-party supplier who then ships the product directly to the customer.

10. Crowdfunding: Funding a project or venture by raising small amounts of money from a large number of people, typically via the internet. Platforms like Kickstarter and Indiegogo are popular for this.

11. Consulting and Coaching: Offering expert advice or coaching services in a specific field, often conducted entirely online via video calls, email, or webinars.

12. Cloud Services: Beyond APIs, this includes data storage, cloud computing, and various IT services provided over the internet.

13. Software as a Service (SaaS): Offering software applications over the internet, typically on a subscription basis. This includes platforms like Salesforce, Slack, and Zoom.

14. Artificial Intelligence as a Service (AIaaS): Offering AI services over the internet, typically on a subscription basis. This includes platforms like Google Cloud AI, Microsoft Azure AI, and Amazon AI.

15. Platform as a Service (PaaS): Offering a platform for developing and deploying software applications over the internet, typically on a subscription basis. This includes platforms like Heroku, AWS Elastic Beanstalk, and Google App Engine.

16. Infrastructure as a Service (IaaS): Offering cloud-based services for computing, storage, networking, and security over the internet, typically on a subscription basis. This includes platforms like AWS, Microsoft Azure, and Google Cloud.

17. Blockchain as a Service (BaaS): Offering blockchain services over the internet, typically on a subscription basis. This includes platforms like Microsoft Azure Blockchain, IBM Blockchain Platform, and Amazon Managed Blockchain.

18. Internet of Things as a Service (IoTaaS): Offering IoT services over the internet, typically on a subscription basis. This includes platforms like Microsoft Azure IoT, AWS IoT, and Google Cloud IoT.

19. Robotic Process Automation as a Service (RPAaaS): Offering RPA services over the internet, typically on a subscription basis. This includes platforms like Microsoft Power Automate, AWS RoboMaker, and Google Cloud RPA.

20. Business Process as a Service (BPaaS): Offering business process services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365, AWS Connect, and Google Cloud Contact Center AI.

21. Human Resources as a Service (HRaaS): Offering HR services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365 Human Resources, AWS WorkDocs, and Google Cloud Talent Solution.

22. Marketing as a Service (MaaS): Offering marketing services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365 Marketing, AWS Pinpoint, and Google Cloud Marketing Platform.

23. Sales as a Service (SaaS): Offering sales services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365 Sales, AWS Marketplace, and Google Cloud Sales Platform.

24. Finance as a Service (FaaS): Offering finance services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365 Finance, AWS Financial Services, and Google Cloud Financial Services.

25. Accounting as a Service (AaaS): Offering accounting services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365 Finance, AWS Financial Services, and Google Cloud Financial Services.

26. Legal as a Service (LaaS): Offering legal services over the internet, typically on a subscription basis. This includes platforms like Microsoft Dynamics 365 Legal, AWS Legal Services, and Google Cloud Legal Services.

Some of these models overlap with each other. For example, SaaS, AIaaS, PaaS, IaaS, BaaS, IoTaaS, RPAaaS, BPaaS, HRaaS, MaaS, SaaS, FaaS, AaaS, and LaaS are all examples of cloud services. However, I'm going to treat them as separate models for the purposes of this project.

Generating a more specific business model

After choosing a general business model, I'll use it along with other criteria to generate a list of specific business models that I could potentially pursue.

I'll engineer a prompt to guide AI to generate business model options based on certain criteria that I'll define. From the generated list, I'll then choose one of the business models to refine and test.

I'll be choosing the business model based on additional criteria that will be used to rank the options. The criteria will be largely based on the following questions:

1. Can the product be created using AI?

2. Could the product be created in under 90 days?

3. Is the product something that can be sold online?

4. Is the product something that can be sold as a subscription?

5. Is there evidence that there is a market for the product?

6. Is there evidence that people are willing to pay for the product?

7. Could the market support a new $1M+ business?

8. Is there a lot of competition?

9. Is there a clear path to profitability?

10. Is the market likely to grow in the future?

11. Are there significant barriers to entry?

12. Is there a clear path to determining product-market fit?

13. Could product-market fit be evaluated in under 6 months?

14. Can the model be fully automated and scaled using AI?

15. Could the model be replicated with other products?

Answers to some of the questions will eliminate certain business models. For example, if the product can't be created using AI, then it's not a good fit for this project. Or, if the product can't be created in under 90 days, then it's not a good fit for this project.

The answers to other questions will be used to rank the remaining options. For example, if there is a lot of competition, that doesn't necessarily disqualify the model, but it would rank lower than others that have less competition.

The ranking and scoring algorithm will initially be based on my own intuition and experience. However, I will be using AI to help me evaluate the ideas and improve the ranking algorithm over time.

I'll also use other proven tools and methodologies to help me evaluate different models. For example, I'll be using the Business Model Canvas and the Lean Startup methodology. These frameworks will also be used to help guide the AI training or prompting that I'll be using to improve the ranking and evaluation process.

Step 2: Decide what to sell

The specific business model won't necessarily define exactly what will be sold. For example, if I choose the e-commerce model, and I decide to sell data sets, then I'll need to decide on what specific data sets to sell. Deciding on the specific product will largely be based on the following questions:

1. Could the product or an MVP (minimum viable product) (MVP) be created in under 90 days?

2. Could the product be created using AI?

3. Could I generate revenue from the product within 90 days?

4. Could I decide to scale or bail within 6 months?

Step 3: Define the target audience

Ideally, I'll be able to adapt the product for multiple audiences. However, I'll need to start with a specific audience to target. This decision will largely be based on the following questions:

1. Is there a specific audience that I'm already a part of?

2. Is there a specific audience that I'm already familiar with?

3. Is there a clear path to reaching the audience?

4. How much would it cost to reach the audience?

5. What is the size of the audience?

6. Is the audience likely to grow in the future?

7. How quickly could I test the product with the audience?

8. What do I expect the customer acquisition cost (CAC) to be?

9. What do I expect the lifetime value (LTV) of the customer to be?

10. What do I expect the customer retention rate to be?

11. What do I expect customer retention cost to be?

12. What do I expect the customer churn rate to be?

13. How confident am I in my answers to the above questions?

To be continued…

I’ll pick up in later posts to discuss steps 4 - 10. But in the next post, I’ll detail how I’m generating the business model to align with my values, interests, areas of knowledge, and my key objectives. I’ll discuss the AI tools I’m using and provide examples of the prompts I’ve come up with. So stay tuned!

Subscribe now

Hi, I’m Steve Tingiris, welcome to the Dabble.AI project. I’ll be the guy behind all the posts and the photographer of all the photos you’ll see here. My goals for this project are simple. I want to figure out how to leverage AI to create and grow a profitable business. I want do that using only AI tools - by myself - without any employees or contractors. It believe it can be done. But I haven’t done it yet. I know a lot about AI and starting profitable companies. So I feel as qualified as I could be. Making this happen is also personal to me.

My current company Dabble Lab helps large enterprises leverage AI to automate the things that it no longer make sense for people to do. This obviously has an impact on the lives of the those whose jobs get automated. But ironically, there is a strong likelihood that the type of work we do (writing code, training ML models, etc) will soon be eliminated as well. However, I don’t see that as a bad thing. Because I believe AI has far more potential for lifting people up than pushing them out. That’s what I’m trying to prove here.

I’m committed to getting one or more profitable businesses up and running and documenting the process. I’m also committed to creating a roadmap for anyone else interested in doing something similar. Hopefully that will include my team members at Dabble Lab, along with friends and family members. Assuming I can figure this out of course.

I’m borderline obsessive about documenting things. So, you’ll have access to plenty of details - if you want them. But I’ll use my posts here mostly as progress updates. All of the details will be available on GitHub at: https://github.com/dabbleai. I might also publish some video tutorials.

I’ve been thinking about this project for some time - since the release of my book Exploring GPT-3 in 2021. Yes, that was a plug. But not in hopes you’ll buy the book. It’s super outdated now. So, unless you’re into GPT history, don’t buy it. The book plug is more about establishing credibility. There is a ton of “get rich quick with AI” stuff out there. This isn’t one of those. I don’t have things figure out yet. I just have a decent amount of experience and clear objectives.

Objectives

1. My primary objective is to figure out how to leverage AI to plan, build, launch, and grow a profitable company, that has only one employee - me.

2. I want to do this without spending a lot of time or money. I'm not sure how much time or money I'll spend. But I'm going to try to spend as little as possible.

3. I also want to create a blueprint or at least a general roadmap that others could benefit from.

I've been doing a lot of research, planning, and dabbling over the past few months. Now I have a general idea about how I'm going to approach this. But that’s as far as I am at the moment. I have not even decided on a business model. I'm going to use AI to help me with that. I'll share more about the general approach I’m going to take tomorrow.

Also, as I move forward I'll provide regular updates on my progress. Including:

• What I'm doing

• The tools I’m using

• How much time I'm spending

• How much money I'm spending

• What's working

• What's not working

• My profits and losses

I hope you'll subscribe and join me as I try to figure this out.

Subscribe now