How Blockchains Can Make AI Safer

For those who’re inquisitive about rising tech, it’s more durable to think about a cooler job than Allison Duettmann’s. She’s the CEO of the Foresight Institute, a suppose tank based in 1986 that’s dedicated to advancing science and tech that’s arguably undervalued — fields like nanotechnology, area exploration, and harnessing cryptography to construct safer AI.

Allison Duettmann is speaker on the AI Stage at Consensus 2024, in Austin, Texas, Might 29-31.

Duettmann lives on this planet of Huge Future Concepts. She’s the coauthor of “Gaming the Future: Applied sciences for Clever Voluntary Cooperation”; she’s the founding father of Existential Hope (with a mission of “Considering Huge for Optimistic Futures”); and he or she’ll be becoming a member of Consensus 2024 on the inaugural AI Summit to share insights on what’s legitimately helpful on this planet of AI+Crypto, and what’s simply hype.

In a current Zoom interview, we cowl the staggering upside of nanotechnology, dig into how crypto might assist the event of AI, and we could have by chance simply birthed a billion-dollar token.

Interview has been condensed and flippantly edited for readability.Let’s begin with a curveball. With some tech that appears essential to you, however doesn’t get numerous press: Nanotechnology. If we crack the code of nanotech, what’s an instance of a use case? How would the world be completely different?Allison Duettmann: Mainly, it might revolutionize all the means wherein we produce issues. So, sure, particular person functions are thrilling. However extra usually, it’s nearly like we’re nonetheless producing issues on this tremendous stone-agey means. We take the sources we now have and attempt to strap them collectively.

But when we might get rather more exact [through nanotech] about what sorts of sources we use within the manufacturing course of… Like, presently, we’re working at a completely improper stage. We’re producing a lot waste within the course of of manufacturing.

Think about a 3D-printing technique, for instance, however on a tiny molecular scale. So that you’d have large features on that entrance. Nanotech might have actually fascinating implications on {hardware} for computing, and actually fascinating functions in longevity expertise.

How so?

For cryonics, for instance, you want nearly molecular nanotechnology for the revival of sufferers — reminiscent of on the drug supply stage or the nanobots in your bloodstream perspective. Then there’s additionally innovation from an environmental perspective; carbon drawdown can be enabled via molecular nanotechnology, or waste removing within the biosphere.

Given the large upside, why hasn’t there been extra growth in nanotech?

One factor that is so fascinating about nanotech is that for anyone particular person utility, there’s most likely a neater approach to do it. But when we had this sort of common function, excessive throughput nanotechnology, then you could possibly be utilizing that for a lot of completely different fields.

So there’s by no means one incentive to develop it for a person actor. That’s why progress has been so gradual. It’s nearly like this public good that after it’s created, it may trigger these revolutions throughout completely different industries. However for any particular person utility, there’s by no means a novel ingenious to develop it.

That’s so fascinating. I say this solely principally in jest… it appears like candidate for a crypto undertaking. As a result of it’s one thing which may take an enormous crowd to capitalize, with the advantages conferred to a bigger neighborhood. Is there a NANO token?!

Properly…

If not, let’s begin one, Allison!

Sure, now that we’ve mentioned it, the Nano DAO.

No, however I believe in all seriousness, it suits very well. There are numerous folks within the Foresight neighborhood which might be additionally in crypto and care a bunch about nano. Quite a lot of the early sci-fi nano concepts match very well with the philosophy of crypto.

Okay let’s speak about crypto and AI. What do you see as probably the most promising, professional ways in which blockchain will help with the event of safer or higher AI?

I suppose I’d say there are three buckets. One can be the safety facet of issues, the place the Web3/crypto area has numerous expertise with failing quick if you do not have safe techniques constructed.

As a result of most of the techniques within the permissionless or immutable areas — in the event you do not construct them accurately — there could also be a million-dollar bounty hooked up to them. So I believe this spirit of constructing security and having an nearly unintentional vetting course of is a vital mindset, as we construct up AI techniques.

And the second?

The second is on the facet of, how can we leverage applied sciences that the crypto area can be engaged on, to construct AI techniques that we’d need from a human perspective?

Are you able to give an instance?

So, there could also be extra privacy-preserving functions which might be doable as we draw extra on the cryptography areas. They could allow us to provide higher functions, for instance, for human well being and in finance, leveraging cryptographic applied sciences. And we might be able to use cryptographic applied sciences to assist us create higher, very superior AI techniques.

And what’s the third and ultimate bucket?

The governance facet of AI. So, you’ll be able to presumably use cryptographic applied sciences to show that you just’re following sure security benchmarks or requirements. Somewhat than having to share the entire data that’s presumably proprietary, you’ll be able to solely share data that’s wanted to show you’re hitting sure security benchmarks.

Nice, and I do know we’ll unpack all of this extra on the AI Summit. Talking of, what are you most wanting ahead to at Consensus this 12 months?

I’m actually wanting ahead to discovering what the state of crypto and AI has actually come to at this level. I come at it from the foundational analysis facet of issues or extra on the form of cryptography and safety facet of issues. However I additionally know that many individuals have been actually constructing arduous within the Web3 and AI area. So, I am actually to see which of my hypotheses — or which of different folks’s hypotheses — have come out. Who’s constructing what proper now? And the place are there doable synergies throughout completely different merchandise? I’m excited to see what’s taking place in that area.

You and me each. See you at Consensus.