Dan Pratl, CEO and founder of Quadron, traces the company to what he calls “a very serendipitous path, starting at the Securities and Exchange Commission in the Great Recession, then into open source, into crowdfunding, and then into crypto.” That path, he told a panel of analysts and journalists at TelcoForge’s recent Forging Ahead online event, led him to one conclusion:
“We need a new systems architecture for the next generation of verification and identity.”
This is not trying to remake existing identity services, but rethinking how we understand identity as a concept; less about who you are and more about what you’re capable of. Pratl’s starting premise is that AI has broken the old signals people used to demonstrate competence.
“Expertise is becoming the new scarce resource in the world… because AI is making polished output incredibly abundant. Anyone can generate a proposal, a memo, a pitch or a strategy that sounds incredibly impressive.”
The question buyers, employers and collaborators need answered has changed: “The question is no longer just can this person produce something that looks good? The question needs to become ‘Can this person be trusted to know what they are talking about?'”
He frames this as the next phase in the history of online identity. Captcha solved “are you a real human?”. Passwords and two-factor authentication solved “do you control this account?”. Neither answers what he says the market needs now: “An agent, an employer, a customer or a counterparty will need to know something more, deeper. What is this person actually trusted to know, decide and do?”
As a result, Pratl is building his company Quadron as a way to verify and authenticate human competence.
“Quadron turns that evidence into a living system of record for expertise,” Pratl explains, by creating verified records of capabilities and decisions on a decentralised blockchain.
While that sounds simple, the reality is that it’s not. Like people buffing up their LinkedIn bios, this could be open to abuse. “A bad claim on chain is still a very bad claim. The harder problem’s actually upstream, creating records that are worth credentialing in the first place.”
We’ll come back to that one.
A Man’s Worth
Prakash Sangam of Tantra Analyst pressed for the mechanics. “You’re suggesting a third-party system that the systems will go check for – instead of credentials, the expertise – and presumably you will manage that system. And for every query you provide a response on whether that person’s expertise is valid or not. Is that what you’re thinking?”
The short answer is yes, but the reason behind it was thought-provoking.
“We need a new economic framework for evaluating human ingenuity and rewarding it, outside of the intellectual property system today. And we need a new technological framework for deploying agents based on verifiable work that human beings actually are standing behind.”
This is really the fundamental issue. We’re at a point where AI agents are able to take action rapidly and to a great degree autonomously. However, they cannot be prosecuted for mistakes, have their own bank accounts, or be rewarded for successes in any meaningful way. Ultimately there has to be an entity behind them. And, while an agent is essentially just code, intellectual property regulations are not well suited to claiming, managing and rewarding things that may be created and gone in a day.
So there’s an upside to being able to prove that you know what you’re doing when it comes to creating or managing agents, but it’s the same when it comes to being the ‘human in the loop’ whose judgment prevents agents from ruining a company’s reputation or handing over sensitive information. We can make whatever claims we like today, and the best rewards go to the biggest posers rather than the people with the best professional expertise and judgment.
While this is the upside for individuals, it also answers an emerging problem for employers as well. “A lot of individuals, in my personal experience, view these new institutional AI solutions that aren’t ChatGPT and Claude as either surveillance or training your replacement… You’re not going to get the highest quality work from people, because they are very concerned with being replaced.”
Now? Even if you think you may be getting replaced, there’s a real incentive to perform well because you can take that performance with you elsewhere and get the recognition for it.
Expert Gits
Asked again later whether certification would cover AI agents as well as people, Pratl laid out the model directly:
“There needs to be a GitHub for expertise, so to speak. And critical to that is the encoding, and then the redeployability, of verifiable expertise.”
He predicted existing professional networks won’t survive the shift: “The platforms we see today, where you go to someplace to meet other people, are largely going to go the way of dinosaurs, as will the advertising model.” In their place: “a very peer-to-peer, me-to-you transfer of value, where agents become the conduit, while the intelligence lives in network to enable those agents to do very discrete and specific things.”
Pratl made the argument that we should look at markets as ways to signal what people believe – in particular, prediction markets reflect what people believe enough to put money on. He connected this to a long-running gap in open source models:
“The development framework of the past 35 years never had a remuneration structure. It was always obligation or fascination that maintained these repos and we saw those failures time and again. If we create some type of reputation-based compensation structure using speculation, all of a sudden you can fill that gap.”
Where does that lead, though?
“I view a time where this thin client right here [on my phone] has a trusted execution environment on hardware where we’re actually running small models at the edge. Good enough is good enough from an agent perspective, from a model perspective.”
System of Control
Telecoms.com editorial director Scott Bicheno summarised what Quadron stood for, and also the key concern with it. “It’s like Google page rank, but for people.”
“If there was a central point of control, that would be a ridiculous amount of power to have in the hands of whoever had that control. The fact that someone had that power would fundamentally undermine the trust that it’s designed to create.”
What’s more, it could become “some sort of dystopian social credit system type of thing, where you end up having punitive de-ranking for some fairly accidental or trivial thing.” Sangam added that PageRank itself is “managed by one company with unknown algorithm, which supposedly is changing very frequently as well.”
Pratl agreed with the stakes. “I think social credit scores are inevitable. It’s a question of how we do them productively.”
He based his argument on established precedents. “Academic citations already exist, patent citations of other patents already exist… It’s already there.”
As to who should govern it, he left the question open: “Do we use standards bodies, or do we adjudicate this in the courts like actual legislation?… this is much more broad in terms of its societal impact.”
Wrangling Truth From Power
Sangam asked how disputes get resolved without a central body to appeal to.
“I think it’s three layers,” said Pratl.
“One is person to person, peer to peer… Using a programmable token means that if there is identified, rules-based malfeasance you can actually have negative consequences.
“Second is the same thing GitHub does when something bad happens; remediation, on-platform remediation.”
The third: “Arbitration in more of a formal context. And I don’t think as an infrastructure solution we want to be in that… This is so critical – we’re talking about our livelihoods and our value here – I think that does become societal at that level.”
How could a decentralized, disintermediating model coexist with governments seeking sovereignty over their citizens’ data? Pratl treated friction with regulators as expected: “disintermediation was the whole point of crypto,” and placed the resolution outside Quadron’s hands.
“All that starts at the regulation side and with the governments. And I think that’s inevitable because the impact is going to be at that societal level.”
Sangam drew the direct comparison: “This is going to have the same headwinds that crypto has right now… because it is wrestling the control away from the central governments, and governments are not good at giving away control.”
Pratl argued regulatory postures would likely shift rather than block this approach, citing his own background: “I was at the SEC, I know the FTC quite well. The digital utility token exemption from the SEC is the pathway.”
It’s important to note that this token is being discussed but isn’t yet encoded in any law or process.
He linked this to a broader prediction: “Our Social Security system in the United States starts to fail as Universal Basic Income becomes more of a proposal, that gets… greater support over the next 10 to 15 years. I think something that looks like a passive annuity structure based on actual work, verified progress is going to get a lot more support.”
Having said that “I understand the SEC and the CFTC and others really stand in opposition of this stuff today.”
That’s perhaps an understatement. Control lies in the hands of governments and the very wealthy today, and it’s going to take a lot to change that.
However, Pratl seems to be taking a Robin Hood view of how decentralisation can work, and ultimately all it needs is a shared agreement that it does – exactly what the current financial system operates on. This could change what ‘value’ and ‘worth’ mean, as he summarised:
“Expertise is the hidden, invisible value that actually supports all of the economic systems we have today. We’ve just never had the systems to capture and encode it, then evaluate it and monetise it effectively.”
