"Fortune 500s ask just one thing before touching blockchain: ‘What’s my..."
From decentralized identity to AI-proof trust, Alex Apeldoorn has helped shape the infrastructure behind Web3’s most ambitious layers. In this conversation, Alex shares what’s broken in Web3 UX, why deep-fakes could be AI’s biggest threat, and how blockchain offers a surprising fix.
From Blockpass and self-sovereign identity to now driving L1 and L2 strategy at Cardano Foundation, what’s been your North Star when it comes to finding product-market fit in Web3?
A core tenet for me is to allow the most amount of people to have the highest degree of self-determination without hindrance. Throughout my career, this has taken many different forms, but the tenet remains the same. How do we make it easy to interact with Dapps, easy to secure your private keys, easy to do trustless commerce, proof something without disclosure and bring the tools to make this to the builders that need them? Technologies like social recovery and MPC, ZKP and DID, data availability, compartmentalisation, etc are the tools for entrepreneurs and innovators to build products for people to gain that self-determination.
You’ve worked across some of the most complex layers of blockchain, ZKPs, DIDs, DEX aggregation, where do you think the real “unlock” for mainstream Web3 adoption lies right now?
Over the last decade, the crypto industry has shown remarkable levels of innovation and growth, solving evermore complex hurdles that were not deemed possible before. From cryptography to networking, data layer orchestration and provable software execution. With all these tools and innovations, I think the next renaissance will come by hiding the technicalities and working towards a Web2 experience in a Web3 world. The real challenge is creating design patterns consumers know from Web2 while giving the users maximum authority over their personal data and assets.
One of the design studies we conducted was on the typical adoption process of a Dapp. On average, in Web3, people spend 12–18 hours researching a Dapp before using it. Compare this with less than 5 minutes in the Web2 world, this gives a clear indication that we need to do better for our users. With the latest innovations, there is really no reason why we can give Web3 users a Web3 experience.
You've helped Fortune 500s ease into decentralized tech. What's the biggest mindset shift you've seen in those conversations, and where are they still getting stuck?
Every conversation I’ve ever had with F500 companies always resulted in the same question: “What is my ROI?” For each organisation, this means different things. For some, this means offboarding access to personal data to shield them from GDPR; for some, it's international remittance; and for others, it's how to internationally collaborate and share data without disclosing information. Whatever it is for companies this size, it's always better to measure twice and cut once when designing a solution.
With your focus on user journey mapping and A/B testing, is there something counterintuitive you’ve learned about how users engage with Web3 products?
My most counterintuitive observation over the years is that no matter how much of an expert you are in your field, you will never know how your product is actually used. An old engineering adage goes: “No matter what tool you design, assume it's also going to be used as a hammer” which usually holds true. To really understand the user, a good product manager needs to assume they know less than the user and ask the counterintuitive questions.
An example that comes to mind is when we released a desktop wallet that became significantly more popular than the browser extension or the mobile wallet, so we started to investigate. Turned out that people could use the local terminal to run bash scripts against the wallet and even build dapps on top of the wallet. It allowed them to customise their experience and use the wallet in ways we had never thought of.
How are you currently exploring AI’s potential, whether in product discovery, user education, or even governance? Any standout experiments or learnings?
AI is an excellent example of a tool that is being used in ways we never imagined at the start. What we are seeing is that there is an explosion of creation, and these creations are being fed back into the LLMs. By current estimate, there is more AI content than human-made content and it's massively distorting the accuracy of the LLM output. LLMs are now being trained on pictures of people with 6-fingers and over time these errors will propagate through the system resulting in higher error rates.
A second more nefarious use of AI is the deep-fakes that are created to get access to systems and corporations at a scale unheard of. Hackers are getting access to internal systems through helpdesks, colleagues that are being impersonated, or counterparties that have been affected. The current solution is to make evermore complex AI tools to fight these deep-fakes, but I think a more fundamental approach needs to be taken.
Both of these problems can be solved using blockchain and DIDs as a root of truth, but this requires the rethinking of data and data ownership. It allows LMMs to recognise original content from AI generated content and it will allow for companies to recognise deep-fakes with real people. The next step for AI is going to be a public permissionless blockchain and a collaboration between tech and creatives.
Solidity Challenge 🕵️♂️
The Tree Bard wants to find the maximum depth of his verse structure. What's the flaw in his poetic function?
Solidity Challenge Answer ✅
The base case should check if the root is null, not if its value is 0, as 0 could be a valid node value.
