Baskadia

My Journey Through HackQuest India × Bittensor Co-Learning Camp

When I first came across Bittensor, I understood it mainly as a project connecting Artificial Intelligence and blockchain. But the more I explored it during the HackQuest India × Bittensor Co-Learning Camp, the more I realized that Bittensor is not simply about putting AI on a blockchain.

The interesting part is the network that forms around AI: miners, validators, subnets, incentives, competition, collaboration, and decentralized intelligence.

This Co-Learning Camp gave me an opportunity to move beyond just reading about the ecosystem. Through the townhalls, discussions, learning sessions, and hands-on mining quest, I was able to understand how the different components of Bittensor fit together.

More importantly, the experience changed the way I think about decentralized AI.

Instead of asking only, "How can AI models become more powerful?", I started thinking about another question:

How can an open network coordinate people and machines to continuously improve AI services through incentives?

That question became one of the biggest takeaways from my Bittensor learning journey.

What Exactly Is Bittensor?

At a high level, Bittensor is a decentralized network designed around the idea of creating an open marketplace for machine intelligence.

Traditional AI development is usually concentrated around organizations that own the infrastructure, models, data, and deployment systems.

Bittensor takes a different approach.

It creates a network where different participants can contribute computational resources, models, AI services, or other forms of intelligence and receive incentives based on their contribution.

The ecosystem uses blockchain-based mechanisms to coordinate these participants.

This is where concepts such as TAO, subnets, miners, and validators become important.

One of the things I found particularly interesting was that Bittensor doesn't treat "AI" as a single service.

Instead, different AI-related problems can exist within different subnets.

This creates a more modular ecosystem where specialized networks can focus on specific intelligence-related tasks.

Understanding the Bittensor Ecosystem

During the camp, I gradually started connecting the different pieces of the ecosystem.

1. TAO

TAO is the native token associated with the Bittensor network.

It plays an important role in the network's incentive mechanism and economy.

But what interested me more was understanding why an incentive token is necessary in the first place.

If participants are contributing computation, models, predictions, or other resources, the network needs a mechanism to reward useful contributions.

That is where token-based incentives become part of the architecture.

2. Subnets

One of the most important concepts I learned was the idea of subnets.

A subnet can be thought of as a specialized environment within the Bittensor ecosystem where participants work toward a particular intelligence-related objective.

Instead of having one network where everyone performs exactly the same task, Bittensor allows different subnets to focus on different problems.

This opens up possibilities for specialized AI services.

For a beginner, I found this analogy helpful:

Bittensor is like a large ecosystem, while subnets are specialized communities inside that ecosystem.

Each subnet can have its own rules, participants, incentives, and purpose.

Understanding this concept made the overall Bittensor architecture much easier for me to visualize.

Miners and Validators: The Two Roles That Changed My Understanding

Before joining the camp, I had encountered the terms "miner" and "validator" in blockchain ecosystems, but I had not fully understood how their roles could work together in a decentralized AI network.

The Bittensor ecosystem gave me a different perspective.

Miners

Miners provide the actual service or intelligence required by a subnet.

Depending on the subnet, this could involve things such as:

• Running models
  

• Providing inference
  

• Performing computations
  

• Generating responses
  

• Supplying data or other useful outputs
  

• Competing based on the quality of their contribution
  

The important point is that miners aren't simply rewarded for being online.

Their contribution needs to be useful according to the subnet's evaluation mechanism.

Validators

Validators evaluate the contributions made by miners.

They interact with miners, assess their outputs, and assign weights based on performance.

This creates an incentive loop:

Miners contribute → Validators evaluate → Useful contributions receive higher weight → Incentives influence future participation

That mechanism was one of the most interesting parts of Bittensor for me.

It resembles a decentralized feedback system.

The network isn't simply asking:

"Who has a computer running?"

Instead, it is trying to determine:

"Who is providing valuable intelligence according to the rules of this subnet?"

My Hands-On Mining Experience

The theoretical concepts became much more meaningful when I participated in the Testnet Miner Quest.

This was probably the part of the camp where I learned the most.

Reading about mining and actually trying to set up a miner are two completely different experiences.

During the hands-on process, I had to understand concepts such as:

• Wallet
  

• Hotkey
  

• Coldkey
  

• Network
  

• NetUID
  

• Miner UID
  

• Subnet
  

• Registration
  

• Miner configuration
  

Initially, these terms can feel overwhelming.

But once I started connecting them together, the process became much easier to understand.

Setting Up the Wallet

The first important step was understanding the wallet structure.

The wallet isn't just an ordinary login mechanism.

It plays an important role in interacting with the Bittensor network.

Understanding the separation between different keys also made me appreciate why blockchain systems use cryptographic identities rather than conventional usernames and passwords.

For someone coming from a traditional software development background, this was an interesting shift in thinking.

Understanding Hotkeys and Coldkeys

The distinction between hotkeys and coldkeys was another concept that I had to pay attention to.

A simplified way of understanding them is:

Coldkey → associated with ownership and security

Hotkey → used for operational activities

This separation is important because operational activities can require frequent interaction with the network, while highly sensitive ownership information should be protected carefully.

This was also a reminder that learning Web3 isn't only about understanding commands.

Security practices are equally important.

The Miner Quest

The mining quest was where the learning became practical.

Instead of simply watching a demonstration, I had to work through the process myself.

I learned about identifying the relevant subnet using its NetUID, configuring the miner, and understanding how a miner becomes part of the network.

One of the most satisfying moments was seeing the setup move from being just a collection of commands and configuration steps to something that actually represented participation in the network.

The process also taught me an important lesson:

Hands-on experimentation exposes gaps in understanding that tutorials can easily hide.

While reading documentation, it is easy to feel that you understand a concept.

But when you actually run the commands and encounter an error, you immediately discover what you don't understand yet.

That made the mining quest particularly valuable for me.

What Challenged Me?

The biggest challenge wasn't a single command.

It was understanding how all the components fit together.

At first, I encountered terms independently:

Wallet → Hotkey → NetUID → UID → Miner → Validator → Subnet

It was easy to memorize them individually but harder to understand their relationship.

Eventually, I started looking at the ecosystem as a flow rather than a list of terms.

A simplified mental model became:

Bittensor Network

↓

Subnets

↓

Miners provide intelligence

↓

Validators evaluate miners

↓

Weights determine relative performance

↓

Incentives encourage useful contributions

This mental model made the architecture much easier to understand.

What Surprised Me About Bittensor?

The biggest surprise for me was the idea that incentives can be used to coordinate intelligence.

In traditional AI development, improvement is usually driven by an organization.

A company builds the model, manages the infrastructure, decides what to optimize, and controls deployment.

Bittensor explores a different model.

Participants can independently contribute resources and compete within an open network.

That creates an interesting question:

Can decentralized incentives create useful AI services without requiring a single organization to control every component?

I don't think that question has a simple answer.

But I found the experiment itself fascinating.

Bittensor Made Me Think Differently About AI

As someone learning AI and machine learning, I usually think about:

• Models
  

• Datasets
  

• Training
  

• Evaluation
  

• Inference
  

• Accuracy
  

• Deployment
  

Bittensor introduced another layer:

Incentives.

A model can be technically impressive, but in a decentralized network, we also need mechanisms to determine whether its contribution is valuable.

That means evaluation becomes extremely important.

This made me realize that the future of AI infrastructure may involve more than simply developing better models.

We may also need better systems for:

Evaluation + Incentives + Coordination + Open Participation

What I Would Tell Someone Starting Bittensor Today

If someone is completely new to Bittensor, I would recommend not trying to understand everything at once.

Start with the ecosystem's basic vocabulary.

Step 1: Understand the big picture

First understand:

What is Bittensor trying to achieve?

Don't start with commands.

Start with the purpose.Step 2: Learn the four core concepts

Focus first on:

1. Bittensor
   

2. Subnets
   

3. Miners
   

4. Validators
   

Once these are clear, concepts such as TAO, weights, registration, UIDs, and incentives become easier to understand.