Module 11: Sector and Narrative Analysis
AI and Crypto
Chapter 8 of 13 · about 4 minutes
8.1 Two hot technologies at once
Every time a major artificial intelligence breakthrough made the news, a wave of crypto tokens with "AI" in their names rallied. Some were serious projects. Some had added "AI" to their branding the week before.
"Is AI crypto real, or just two buzzwords stuck together?" asked Node.
"Both exist," said Hash. "There are genuine reasons AI and blockchains might work together, and there's a lot of opportunistic branding. This sector is a great place to practise separating the two."
8.2 Where AI and crypto might connect
| Area | The idea | What crypto might add |
|---|---|---|
| Decentralised compute | Marketplaces where people rent out GPUs and computing power | Global, permissionless coordination and payment for compute |
| Data | Networks for collecting, labelling, or selling data for AI training | Incentives and payments for contributors; provenance tracking |
| AI agents with wallets | Software agents that hold crypto, pay for services, and trade | Programmable money that software can use without a bank account |
| Verifiable AI | Proving that an AI model produced a specific output | Cryptographic proofs and on-chain records |
| Proof of personhood | Distinguishing humans from bots online | Identity and verification systems |
| AI-related tokens and communities | Tokens tied to AI themes or AI-generated personalities | Mostly attention and speculation |
8.3 AI agents and crypto wallets
One of the most discussed ideas is that AI agents, software that can take actions on behalf of users, will need ways to pay and receive money. Traditional payment systems require identity checks and bank accounts designed for humans and companies. Crypto wallets and stablecoins can be controlled by software directly.
Possible uses include agents paying for data or computing services, managing DeFi positions, executing trades according to rules, or paying each other for services.
| Promise | Risk |
|---|---|
| Automated payments without human approval each time | An agent with wallet access can make costly mistakes or be manipulated |
| Machine-to-machine commerce | Prompt injection and malicious instructions can redirect funds |
| 24/7 strategy execution | Bugs and poor models can lose money fast |
Safety approaches draw on Module 2: separate wallets for agents, spending limits, permissions restricted to specific actions, smart account controls (Module 2, Chapter 13), and human approval for large transactions.
"So letting an AI hold your keys is like giving a new employee the company credit card," said Mint.
"With limits, approvals, and logs, it can work," said Hash. "Without them, it's a risk."
8.4 Hype versus substance
AI crypto narratives have gone through several boom and bust waves, including sharp rallies and collapses in AI agent tokens. Applying Chapter 2's questions helps:
| Question | Strong project | Weak project |
|---|---|---|
| Is there real usage? | Paying customers for compute, data, or services | Only a token and a roadmap |
| Does it need crypto? | Decentralisation or crypto payments solve a real problem | Would work better as a normal company |
| Does the token capture value? | Token needed for payments, staking, or receives fees | Token unrelated to the product |
| Can it compete? | Competitive on price or features with centralised alternatives | Far behind major cloud and AI providers |
8.5 The competition problem
Decentralised AI projects compete with some of the largest and best-funded companies in the world, which operate enormous data centres and research teams. To succeed, decentralised alternatives usually need a clear advantage: lower cost through spare capacity, censorship resistance, privacy, open access, or a niche that centralised providers don't serve.
Without such an advantage, the narrative may attract investment while the product struggles to find users.
8.6 Sector signals
| Signal | What it suggests |
|---|---|
| Revenue from real customers growing | Genuine adoption |
| Compute or data marketplace usage rising | Product traction |
| Agent transaction volume rising, excluding test activity | Agent payments becoming real |
| Tokens rallying on outside AI news without product changes | Narrative-driven moves (Chapter 2) |
| Many projects rebranding to AI | Possible mania stage |
8.7 Try it yourself
Choose three AI-related crypto projects in different sub-areas from section 8.2. For each, find: what the product does, whether it has paying users, how the token is used, and how it compares with a centralised alternative. Then check how their token prices moved around a recent major AI news event outside crypto.
Key takeaways from this chapter
- AI and crypto may connect through decentralised compute, data, AI agents with wallets, verifiable AI, and proof of personhood.
- AI agents can use crypto wallets and stablecoins to pay for services, but need strict limits and controls.
- AI crypto tokens have gone through sharp boom and bust waves, often driven by outside AI news.
- Test projects on real usage, whether crypto is needed, token value capture, and competitiveness.
- Decentralised AI competes with very large centralised companies and needs a clear advantage.
- Watch revenue, marketplace usage, and agent activity, and beware of rebranding waves.