This week, I tried to explain Bittensor to my unsuspecting wife. Her response? (after I promised I wouldn't spend an hour trying to explain it):
"So you're basically saying a bunch of computers argue about who’s the smartest and get paid for it?”
Not a bad analogy, and I prefer it to the usual Bitcoin for AI argument.
But something important to note here is: you don't need an Nvidia powered GPU farm to participate in this ecosystem.
The Funnel Framework
As someone knee deep in Python tutorials and Kaggle datasets, I’ve been inspired by Travis the TAO Templar’s YouTube video exploring ways to make money on Bittensor.
There are so many ways to contribute to the Bittensor ecosystem, and I think something that demonstrates this well is very similar to a marketing funnel.
The top of funnel items are most easily accessible (but not always easiest to execute), moving down to bottom of funnel items for those with extremely specialized skillsets:
Bottom of the Funnel: Miners, Validators & Subnet Owners
Specialized skills required: ML engineering, sysadmin, cryptography.
Miners train AI models, validators evaluate their outputs, and subnet owners build the infrastructure. They’re the ones getting their hands dirty with tensor shapes, bond investments, and GPU meltdowns.
Having only recently committed to this learning path, I've already had a few existential crises thinking how could I ever learn to mine and contribute at this level, these guys are already so advanced and mining is now crazy competitive.
Middle of the Funnel: Traders & Investors
Skills required: Technical analysis, fundamental analysis, risk management.
I’m a numbers guy, not a trader. But my Python skills could eventually let me automate a strategy to scrape subnet price data, detect patterns and trade subnets. Volatility is opportunity and anyone with a data science background could contribute with anything from DTAO arbitrage strategies and data driven analysis to tokenomics modeling.
Investment survival will depend a lot here on feature engineering and data quality, but price is not something I want to focus on in this newsletter, so I will leave this area to others.
Top of the Funnel: Other Contributors
Skills required: Data science, coding, creativity.
Not everyone can be a miner. But subnets need data cleaners, prompt engineers, researchers, and developers. Travis’s story on Maciej Kula proves this, no one paid him to build learnbittensor.org, but his passion made him indispensable as he aspires to educate others on Bittensor.
As everywhere, subnets pay for useful work. If I can structure my data projects to mirror their data pipelines, I can become a candidate for grants or collaborations.
Top of the Funnel: Content Creators
Skills required: Writing, teaching, storytelling.
Travis’s channel and other content creators are the network’s marketers. Subnet owners need someone to explain Bittensor to the confused masses (like my wife, who still thinks TAO is a martial art).
There is even a subnet dedicated to incentivizing this in Bitcast (subnet 93).
This newsletter is my Bitcast application. I’m not filming TikToks (yet), but I’m:
Documenting my Python struggles in public
Providing insights from a beginner's viewpoint interacting with the Bittensor ecosystem
Introducing real-world applications of applying learnings to decentralized AI
Why This Works
Bittensor isn’t a monolith. It is a funnel:
Top: Content creators, researchers and data scientists (like me) enter with low barriers.
Middle: Traders and alpha investors use technical skills to profit.
Bottom: Miners and subnet owners with niche expertise drive the network.
The beauty? Everyone can move through the layers. Travis started as a trader, became a miner, now runs a YouTube channel. Maciej went from basement coder to landing a job at Latent Holdings.
The TAO of Networking
The worst thing I could do now is wait until I’m “ready.”
My next steps, to keep me accountable:
Post 3x/week in community (e.g. Discord) channels.
Pitch my Substack as a beginner’s guide for anyone interested in exploring this ecosystem.
I want to prove decentralized AI isn’t just a PhD playground. So here’s my experiment:
Phase 1 (Python + Data Science): Use practice datasets to master data prep for subnets like Masa.
Phase 2 (Application of knowledge): Use my skillset to provide subnet-specific solutions, validator support, or protocol contributions.
Phase 3 (Community Leverage): Get hired or link with suitable interested parties because I’ve been ranting about their projects for months.
I don't know how long it will take, but this framework isn’t about shortcuts. It’s about becoming irreplaceable in a network that rewards utility.
If you’re a beginner, focus on the top of the funnel. If you’re a data scientist, dig into alpha research. If you’re a sysadmin, start mining.
Bittensor rewards utility, not credentials.
Your Move:
What tier are you in? Hit reply and tell me where you’re stuck. Let’s fill this funnel together.
P.S. If you’re a subnet owner reading this… yes I may be coming for your GitHub issues. Give me some time to sort out my repo first.
Cheers,
Brian