Jumping the Chasm: How AI Solves Blockchain’s Black Box Problem
Jumping the Chasm: How AI Solves Blockchain’s Black Box Problem
In two of our recent newsletters we’ve discussed the growing intersection of blockchain and AI, in terms of security (read more here). This week, however, we discuss how AI will accelerate blockchain adoption by demystifying the user experience and bridging the educational gap between early innovators and mass adoption.
Holes in the Current Internet Necessitate Blockchain
Regular readers will be familiar with our thesis, that there are five key issues that need to be resolved to make the internet tech stack fit for purpose to move beyond the digitalization of content and communications (“Web2”) to the digitalization of commerce and transactions (“Web3”);
The Internet is insecure.
The Internet lacks a system of identity and trust fit for a digital economy.
The Internet is excessively concentrated creating large single points of failure.
The Internet does not have digital monies with which to transact.
The Internet does not have digital assets with which to record, buy and sell, and hold assets in a digitally native way.
Our book, Blockchain Competitive Advantage, describes our thesis in depth. We believe blockchain technology solves these issues, bridging the gap between the current internet and a truly secure, digital, financial experience. We also knew early on that those companies and projects that solved these five issues would unlock the greatest value creation event the world has ever seen.
But Where Does AI Fit In?
Lately, investors have been asking us where AI fits into the equation. We recently answered this question in depth (read more here), but these were a few of the takeaways…
Will AI accelerate Blockchain adoption? Yes, because AI will take advantage of Internet vulnerabilities and force us to upgrade all databases and our approach to everything we do to make it more secure, with better identity and trust in a way that has less centralized points of failure.
How else will AI accelerate Blockchain? We may also need new human identity solutions so we know what is an AI and what is a human (as Sam Altman and Worldcoin believe).
But What Else Needs Stands in the Way of Mass Adoption for Blockchain?
We’ve already covered the five critical issues with today’s internet technology stack in depth, many times, over the years. But if we could add a sixth, and somewhat more general consideration with regards to adoption, it might be technical complexity.
Simply put, the internet is a complex, global amalgamation of hardware and software. Wittingly or unwittingly, almost every human being on the planet interacts with it daily. But few of us actually understand it. And if it weren’t for the large institutions that now dominate the web (and their investments in training, education, abstraction and UI/UX, which have arguably solved this issue for Web2), it's likely we wouldn’t feel so secure using it.
And in the early days of the internet, many thought it was a fad or a scam. And appropriately so. Interacting with early websites felt scary and foreign, like interacting with a black box. There was little to look at and nobody to ask questions of. It was you versus the screen, and unless you were deep in the industry, it’s unlikely you felt secure putting your personal information out there. And the idea of conducting personal finance on the internet, commonplace today, felt lightyears away.
Maybe some will remember what the websites of today’s tech giants looked like in the late 90s. Here is Amazon, circa 1995…
Blockchain Sometimes Feels like a Black Box
The same can be said of blockchain today. Interacting with decentralized, open source software protocols can be quite jarring. And while the technology is rapidly evolving, the user experience and user interface still has a ways to go. Simply put, for those of us who aren’t blockchain engineers, interacting with many blockchain products is like sending money into the great unknown, hoping it will work.
For example, see a couple screenshots below from the homepage of Curve.Finance, a decentralized finance project which provides a marketplace for the exchange trading of stablecoins, using an automated market maker (or AMM).
To a developer, this might look like business as normal. But to your average citizen, this might as well be in a foreign language. “Gas”.. “Vaults”.. “Pools”.. “Voting escrows”.. It’s enough to confuse most anyone. And even though the technology works and has been proven at some considerable scale – the total value locked (TVL, akin to AUM) on Curve.Finance is currently $1.9bn – few people would feel comfortable interacting with it, given the current user experience.
Furthermore, Curve.Finance is a decentralized autonomous organization, or DAO, meaning it’s an internet-native business collectively owned and managed by its members (or token holders). We won’t dive too deep, but if you wanted to actively participate in the management of the product, like a shareholder voting in a company, this is the interface you’d have to interact with…
Again, if you’re a developer, that’s fine. But if you’re anything short of that, you might be hesitant to put money into it. In short, this is a key barrier to the large-scale adoption of blockchain technology.
But this might more aptly be described as a common technology adoption curve issue, rather than a failure of any tech stack. And the internet has arguably solved this problem already.
The Technology Adoption Curve: Moving from Early to Mass Adoption
Viewed as a technology adoption curve issue, this is a common problem between early innovators and general and mass adoption, often referred to as “The Chasm”.
New technologies often require a revolution in usability to bridge the chasm and evolve from their early adopters and niche hobbyist communities towards mass adoption. As technologies become more practical and more familiar, they move out of the hands of hobbyists and niche communities, into the mainstream.
Blockchain, we believe, is poised for just such an evolution in practicality, navigability and usability. And viewed as a technology adoption curve issue, this is unsurprising.
Early developers and blockchain engineers, like the early architects of the internet, built what made sense to them, not everybody else. And like with the internet, practical tooling and interfacing for the general public came later, setting off a wave of global adoption leading to the internet we know today (and the greatest value creation event to date in modern history).
AI Will Help Blockchain Bridge (or even Jump) the Chasm
At Blockchain Coinvestors, we believe AI and even early Large Language Model (LLM)-based products like ChatGPT will play a critical role in this evolution, allowing blockchain to jump the chasm, in a sense. By providing practical tooling, interfacing and chat-style support and education, we believe AI can shine a light on blockchain’s black box issue and catalyze mass adoption.
LLM-based chat support models, like ChatGPT, will help users understand and navigate blockchain concepts like “hash rates” or “gas fees”, acting as 101 tools, user manuals and support services. (Replacing the large brick and mortar support staff early tech giants employed at the equivalent stage in the internet adoption lifecycle to educate the market, at a fraction of the cost). And AI-based dashboards and analytics will help provide better interfacing and a more practical user experience, demystifying web3 products and insulating users from the backend complexity.
And with over 100m accounts and 1.5bn monthly active users (MAU), ChatGPT has already proven the chat-based model’s popularity. TikTok took nine months and Instagram took nearly two and a half years to reach that point. So we know how potent a basic, conversational AI tool can be. And a survey by Savanta revealed that 47% of respondents have already used ChatGPT for learning purposes, so we have a potential proof of concept for this already.
Just Imagine: A Friendlier End to End Blockchain Experience
Imagine looking for a blockchain-based loan. But rather than navigating a code-heavy interface, like above, to interact with what feels like a black box… you see a simple dashboard with live information. It makes sense at a high level and the buttons are clear.
Maybe you still have some questions so you scroll over to CryptoGPT and ask what is meant by interest in this scenario or how fees are calculated. You get a clear explanation with examples and maybe you ask two or three clarifying questions as needed. With the extra reassurance in hand, you feel confident now to interact with the loan protocol and you take out a $1,000 loan successfully (with no bank or central intermediary involved!).
At year end, you’re still holding the loan and you want to do some reporting or calculate your costs to date for the loan. Rather than pulling raw data from the blockchain and analyzing it yourself, you scroll over to CryptoAnalyticsGPT and ask it for the data you need. You automatically get a clean PDF report of your holdings, transactions, taxes, fees, etc. with easy charts and visualizations. Something you could give straight to your accountant or business partner.
And that’s it. You’re insulated from raw blockchain code by a set of practical tooling, interfaces and support agents which allow you to use Web3 with confidence, regardless of your technical knowledge level.
Blockchain + AI: Jumping the Chasm
That’s the kind of future we envision at the intersection of blockchain and AI in the next few years. Tools like ChatGPT will insulate users from raw blockchain code, adding practical layers of abstraction and shedding light on blockchain’s current black box feel.
And as the potential of AI and models like ChatGPT in the field of technical support and abstraction becomes clear, the chasm faced by prior technologies between early adopters and early majority appears smaller and smaller for blockchain. And the chasm begins to look like a thing of the past.
Author
Christopher Nelson
Head of Digital Asset Research