Starting at Opentensor Foundation
I’m really excited to announce that I have just accepted the role of Director of Developer Relations at the Opentensor Foundation.
For those of you that know me, three of my passions in tech are machine learning, blockchain, and developer advocacy. This is a role that allows me to combine all three of them.
My involvement in developer relations and communities goes back to the early 00s – before “developer relations” existed as a role. I was heavily involved in the Zope web application server, written in Python; and the Plone content management system built on top of it. Communicating via mailing lists, video calls, working via IRC and Github. Organizing development sprints, hackathons, and conferences. The ‘art’ of developer relations didn’t really exist back then. There were no books, no blogs, no nomenclature to define the roles and tasks to be done. It was something that was done by developers, even if they didn’t have the words to describe what exactly it was. Certainly, no companies were hiring for the role of “developer advocate” or “DevRel”.
Fast forward nearly two decades and my introduction to developer advocacy as a formal career was at IBM, working in their London Developer Advocacy team, specifically working on machine learning and AI. I used to run Twitch live coding streams in which I’d explore machine learning and the various Python libraries.
I then went on to take my side interest in the XRP Ledger and move into it as a full-time job, taking on the role of Director of Developer Relations at Ripple. Whereas IBM was a big company with a well-defined DevRel structure and teams across the globe, Ripple was a big-ish company, well established, but with no existing developer relations function in place.
And now, I’ve kinda come full circle. Opentensor is the foundation behind the Bittensor project. They are a young startup, with a small, but rapidly growing team, spread across the globe. Bittensor is a project to build a massive decentralized neural network across the internet. You’ve probably seen some pretty impressive neural networks in recent times, GPT-3, DALL-E, Gato. These are able to write entire articles when given a starting sentence or draw an image from a description. But they all need massive amounts of processing power, thousands of CPUs and GPUs to train them. And the ability to run such large models is limited by end-user hardware and licensing models.
Bittensor is the answer to that. It is a peer-to-peer marketplace that rewards the production of machine intelligence. Imagine if you took Bitcoin’s PoW algorithm and replaced the useless “busy work” done by miners with something genuinely useful? Instead of just hunting for hash values that fit a particular pattern, your computer was contributing to a global shared neural network – each miner is judged by its peers and rewarded based on the quality of new knowledge it is bringing to the network. Ultimately, training up what will become the next generations of AI.
If you told me there was a job in which I could combine AI, blockchain, and DevRel, I’d have said you were just playing LinkedIn recruiter buzzword bingo. Indeed, I’ve seen various attempts to munge AI and blockchain together in the past, but never in a way that seems that it would pass the sniff test with either the blockchain community or the AI community.
But, Bittensor has got my attention.
It is a genuinely legitimate attempt to make machine learning more collaborative. The project was founded four years ago by Jacob Steeves and Ala Shaabana, two machine learning researchers, to create a shared global pool of machine intelligence. A pool that anyone can contribute to (and be paid to do so) and anyone can pay to use. To take the state of the art in AI research and pull it out of the silos it is in and put it on an internet-scale decentralized “brain”. Indeed, much of the architecture of Bittensor follows that of the human brain. You will see references to neurons, dendrites, and axons in the codebase and documentation.
So, over the next few weeks I’m going to be working on our DevRel strategy, and getting my hands dirty starting to work on developing demos, working on documentation, speaking at events, reaching out to developers, and helping out in the community.
You can follow Bittensor on:
For those of you that know me, three of my passions in tech are machine learning, blockchain, and developer advocacy. This is a role that allows me to combine all three of them.
My involvement in developer relations and communities goes back to the early 00s – before “developer relations” existed as a role. I was heavily involved in the Zope web application server, written in Python; and the Plone content management system built on top of it. Communicating via mailing lists, video calls, working via IRC and Github. Organizing development sprints, hackathons, and conferences. The ‘art’ of developer relations didn’t really exist back then. There were no books, no blogs, no nomenclature to define the roles and tasks to be done. It was something that was done by developers, even if they didn’t have the words to describe what exactly it was. Certainly, no companies were hiring for the role of “developer advocate” or “DevRel”.
Fast forward nearly two decades and my introduction to developer advocacy as a formal career was at IBM, working in their London Developer Advocacy team, specifically working on machine learning and AI. I used to run Twitch live coding streams in which I’d explore machine learning and the various Python libraries.
I then went on to take my side interest in the XRP Ledger and move into it as a full-time job, taking on the role of Director of Developer Relations at Ripple. Whereas IBM was a big company with a well-defined DevRel structure and teams across the globe, Ripple was a big-ish company, well established, but with no existing developer relations function in place.
And now, I’ve kinda come full circle. Opentensor is the foundation behind the Bittensor project. They are a young startup, with a small, but rapidly growing team, spread across the globe. Bittensor is a project to build a massive decentralized neural network across the internet. You’ve probably seen some pretty impressive neural networks in recent times, GPT-3, DALL-E, Gato. These are able to write entire articles when given a starting sentence or draw an image from a description. But they all need massive amounts of processing power, thousands of CPUs and GPUs to train them. And the ability to run such large models is limited by end-user hardware and licensing models.
Bittensor is the answer to that. It is a peer-to-peer marketplace that rewards the production of machine intelligence. Imagine if you took Bitcoin’s PoW algorithm and replaced the useless “busy work” done by miners with something genuinely useful? Instead of just hunting for hash values that fit a particular pattern, your computer was contributing to a global shared neural network – each miner is judged by its peers and rewarded based on the quality of new knowledge it is bringing to the network. Ultimately, training up what will become the next generations of AI.
If you told me there was a job in which I could combine AI, blockchain, and DevRel, I’d have said you were just playing LinkedIn recruiter buzzword bingo. Indeed, I’ve seen various attempts to munge AI and blockchain together in the past, but never in a way that seems that it would pass the sniff test with either the blockchain community or the AI community.
But, Bittensor has got my attention.
It is a genuinely legitimate attempt to make machine learning more collaborative. The project was founded four years ago by Jacob Steeves and Ala Shaabana, two machine learning researchers, to create a shared global pool of machine intelligence. A pool that anyone can contribute to (and be paid to do so) and anyone can pay to use. To take the state of the art in AI research and pull it out of the silos it is in and put it on an internet-scale decentralized “brain”. Indeed, much of the architecture of Bittensor follows that of the human brain. You will see references to neurons, dendrites, and axons in the codebase and documentation.
So, over the next few weeks I’m going to be working on our DevRel strategy, and getting my hands dirty starting to work on developing demos, working on documentation, speaking at events, reaching out to developers, and helping out in the community.
You can follow Bittensor on:
- Twitter at: https://twitter.com/bittensor_
- Website: https://bittensor.com/
- Bittensor Discord channel: https://discord.com/invite/3rUr6EcvbB
There was also a fantastic interview a couple of weeks ago of the founders by Jeremie Harris on the Towards Data Science podcast: https://towardsdatascience.com/ai-on-the-blockchain-it-actually-might-just-make-sense-6763389453cc
