I have not programmed actively for years.
I built an app in an hour using the SN1 + SN13 back-ends - it scrapes X for content on a user, then summarises it and provides a response using the Apex LLM.
In a post-product era, the sheer versatility of building on Bittensor will let
Will Squires
2,206 posts
Co-Founder @MacrocosmosAI
- A literal universe of data. 34bn rows, and more coming. Bittensor native test time compute. @OpenAI - @opentensor is coming for you. app.macrocosmos.ai
- go to bed @const_reborn ships EVM wake up three teams already launched integrations (@BlockchatDAO respect) realise I’m not sleeping until dTAO LETS GO.Well … why wait? ✅ Depermissioned EVM Just pushed to Finney.
- What a great time at @endgame_summit - thanks to @madladshad for organising, and all of the TAO family for the supreme alpha. A few thoughts - 🧵
- the disconnect between pace of building on bittensor and complaints about price action is real. just within macrocosmos, in the past two weeks: - we completely rebuilt our front end stack for SN13 (currently undergoing some final testing, but we'll be sharing shortly) - we
- I wrote a short piece on the economics of being a subnet owner in dTAO. I also touch on root weight, and the broader AI landscape. Liquidity, credit, capital, and developments in tokenomics are all critical things we need to focus on as a community.
- Thrilled to announce @crux081208 and I are launching @MacrocosmosAI today - building on Bittensor. We’re building an amazing team, so reach out if you’re passionate about democratising intelligence, and building AI for all humankind. Thanks to @opentensor and @const_reborn
- I have high conviction in @opentensor and @const_reborn - a team that has always focused on building. In dTAO, focus on what people are building - @mogmachine coming through for the ecosystem We're launching Constellation next week - our multi-subnet platform. I can't wait.We are SO EXCITED for dTao. Of course we have made sure you will have the full trading experience for dTao, but it is just one of a host of updates we have for you tonight. .....
- iota is now live. We are starting with a 15B parameter run to test a variety of the mechanisms (overall architecture, dashboard, infrastructure) over 5 layers. The intention is to gather test data from this stage prior to launching a larger competition. We expect CLASP toIt's here: Incentivized Orchestrated Training Architecture... ...live on #Bittensor Subnet 9, IOTA. Our new dashboard is live - featuring: 🎬 Real-time 3D pipeline 🧩 Interactive miner nodes with live data ⚡ Active visualisation of training attributions 🌐 Global network
00:00 - Today, I am incredibly proud to share our white paper outlining our most recent research at Macrocosmos, and our best work to date. Introducing IOTA - Incentivised, Orchestrated Training Architecture, a new paradigm in distributed, internet-scale training, that has the potentialIOTA (Incentivized Orchestration Training Architecture) is a framework for pretraining large language models across a network of heterogeneous, unreliable, permissionless and token incentivized machines. In our technical primer we report the following advances: Incentivized













