Nanoneuro and First Principles

We are building Nanoneuro Systems with first principles thinking in mind:

1. Why should we exist?
2. What makes us feasible?
3. What are our bottlenecks?
4. What would we unlock if those problems were solved?

One thing I’ve learned from reading about Elon Musk and his Tesla and SpaceX stories: the mainstream, the famous, and the high-performing funds did not invest in Tesla or SpaceX. In fact, “thought leaders” and “industry leaders” heavily criticized his ideas.

It didn’t matter. Opinions are opinions.

What mattered was solving the right problems, as Tesla and SpaceX have. Electric cars are now drastically more mainstream with a real climate impact, and SpaceX has single-handedly created a new era of space startups and exploration.

They started with real problems. They ended with real value.

Following this, I thought about Nanoneuro Systems from a first principles approach.

Why should it exist? Many reasons:

1. Because the world is entering into an AI age, and that age demands electricity that the world currently can’t supply, or supplied through dirty fuel.
2. Because the world (surprisingly) is amid a silicon shortage, which will not sustain the growing demand for silicon-based components.
3. Because climate is a growing crisis and a solution needs to be sustainable: not needing millions of gallons of water every year to liquid cool compute centers or the constant mining of raw earth materials.
4. Because our usage of data in AI training will lead to a data-starved future for AI, and that demands new technologies to enable AI training from minimal data in one-shot learning processes.

We’re leveraging new developments in biotechnology and fabrication tech to create a biologics-based computing paradigm that promises 10000x energy efficiencies while being more sustainable. Of course, that means lots of technical challenges we need to overcome:

1. Custom designing, then mass producing a new silicon interfacing base chip.
2. Sustaining brain cells for long-term compute on levels achieved by current silicon architectures.
3. Pioneering methods to allow transfer learning between brain cells to retain learned algorithms.
4. Defining a new biologic-chip architecture with the software stack to enable software developments on it.

But scientifically, these are doable challenges. We’re not even inventing anything new. We’re just integrating existing technologies and adapting them for a unique use-case.

But most importantly, what opportunities would we unlock if our solution works?

We’re building hardware that enables the coexistence of sustainability and AI. We’re building with the future of AI development in mind. We’re building to account for a data-starved digital world. We’re building to bypass future silicon shortages for chip manufacturing. We’re building to reduce global data center energy usage from 20% to 2% in the coming decade. Plus, there has never been a more critical time to rebuild the American chip manufacturing capability as geopolitical tensions rise to levels unprecedented since the Cold War.

And quite frankly, I’ve long sailed past my era of building B2B AI applications. I want to do something revolutionary with a high chance of failing.

And yes, people will disagree with us. Many mainstream, famous, and high-performing funds won’t invest in us because of the impossibilities involved – YC told me to get a PhD before applying with the same idea, or switch to software and be accepted immeadiately.

But know this. There is a world of difference between something that’s impossible and something that’s almost impossible. There will be times when it will seem impossible to us. I already know this – odds are, we are giong to fail. But we will try, like others. The frontier of alternative computing is happening everywhere!

And perhaps one day, there will be a moment where the world’s efforts and moments of defiance will have broken through. One single advancement will change the era. SpaceX has. Who’s next?

We’re enlisting. We’ll try. And we don’t need to convince everyone to believe in us, in what we’re doing. Just the visionaries willing to take a bet on the future¹.

Footnotes

1. I’ve written before that Consensus is Regression to the Mean. In it, I argue that consensus investments in the venture capital space will lead to merely average returns. ↩