Bits, Electrons, & Tokens
projects that are moving the ball forward
Below are a few young projects that portend things to come.
Plural
It’s easy to finance rooftop solar and megasites but surprisingly difficult to do so for projects in between. The former is small and generally covered by consumer credit like financing companies while the latter more effectively spreads the expensive underwriting costs over a larger financial surface area. The bigger the better. The middle is stuck between a rock and a hard place. Plural is using tokenization to collapse the underwriting and fundraising costs for middle market projects in the energy, compute, and general infrastructure space.
USD.AI
Intelligence production is very CapEx heavy. Whether its training or inferencing, running many expensive machines is non-negotiable. Advanced GPUs are extremely valuable but tapping into that value isn’t easy. USDAI is making it easy to tokenize and collateralize a fleet of machines and borrow against them. This unlocks debt financing for non-majors who can’t secure financing from traditional institutions, large and small.
Just the above is impressive in and of itself but from what I can tell the team behind has built it with impressive primitives that can unlock creative financing of a much greater constellation of physical infrastructure outside of advanced GPUs. This looks like just the start of something that could become much bigger.
Zeus
Access to accurate weather predictions has always been important to optimal grid management. Its importance increments up with every increase in variable supply. Sun, wind, rain, and snow all effect demand for and the supply of electricity. Researchers & professional organizations as well as projects like Google’s DeepMind have filled this role. Zeus, is attempting to do the same thing but through a novel mechanism that shows the potential creativity and power of DeAI, and more specifically Bittensor.
Zeus is a “subnet” or sub-network within the Bittensor network. Currently, there are many operating Bittensor subnets, with an upper-limit of 128. Each subnet creator sets the purpose, rules, metrics, and reward system in order to attract contributors that can forward that subnet’s purpose. The Zeus subnet incentivizes the creation of accurate weather predictions. This is an open, co-opetition that anyone or group in the world can contribute to and/or compete within.
The team behind Zeus is reporting impressive predictions that meet or exceed predictions made by the state of the art open source forecasting service Open-Meteo. This is just one of many interesting subnets operating on Bittensor producing digital commodities like weather prediction in completely novel ways.
WOAH!
The following scenario is not too far off from becoming a workable reality:
Imagine 2, 10 MW micro-reactors powering a data center that’s financed by a broad coalition of small financial institutions, funds, and individuals. Lets call it WOAH.
The securities for WOAH are easily accessible via secondary markets on Coinbase, Uniswap, and hundreds of other venues. Holders can also access low cost financial services like collateralized lending on Kraken, Morpho, and many others. In addition, a high performing team contributing to the Zeus subnet is renting rackspace from the data center, providing revenue, some of which filters down to WOAH holders. More importantly they crank out high quality predictions to grid operators all over the world helping them optimally manage their assets. This team is also able to easily collateralize and borrow against their GPUs using USD.AI. This allows them to buy or rent more machines to improve their performance on the Zeus subnet, earning them more rewards.
There is a lot to lament in the world but there are still small, powerful pockets of competence, productivity, and good will. Never black-pill. Do your part, however small, to increase the size and power of those pockets.