January 9, 2025
AI's Hidden Threat of Crypto Centralization
Dear Subscriber,
 |
| By Jurica Dujmovic |
I recently came across a thought-provoking article on Decrypt. It questions whether AI agents, rather than humans, might drive the next wave of crypto adoption.
I’ve talked about AI agents before. To recap, these are AI frameworks that enhance existing models with decision-making capabilities and real-world interactions.
In short, AI agents operate autonomously to do tasks that previously required a human touch.
That’s why I believe they represent the future of AI.
But this piece takes AI agents to the next step. Featuring perspectives from various crypto builders and coinciding with NVIDIA's CES 2025 announcements, it explores the possibility of artificial intelligence becoming the dominant force in blockchain networks.
As someone who straddles both worlds, I have some thoughts on the ideas presented … and the implications should they become reality.
Let’s start with the article’s central bold prediction from an ecosystem lead at the EVM-compatible Layer-1, Monad, known by his X handle, Evan.
He claims, "We will see a billion agents on-chain before we see a billion humans on-chain. Crypto has inherently unfriendly UX for human users. Agents do not care about this friction."
Evan makes some compelling points in his detailed thread. He covers the evolution of AI agents in crypto from frameworks to dApps to DAOs.
The article then points to platforms like Virtuals Protocol (VIRTUAL, Not Yet Rated), a decentralized platform that lets users without technical expertise create and monetize AI agents. With $60 million in revenue and AI agents like Luna (a virtual K-pop star) and aixbt (a purple Pepe frog that provides laconic market insights), Virtuals has certainly captured the crypto zeitgeist.
Its VIRTUAL token powers its ecosystem where AI agents are tokenized, owned by the community and generate revenue through "inference costs."
With projects like Virtuals rolling out, it’s easy to see Evan’s point.
But I have a different take on this vision.
I agree that AI agents will undoubtedly play a crucial role in crypto's future. But to me, the idea that they’ll outpace human adoption feels like putting the cart before the horse.
It's not that AI agents can't or won't participate in blockchain networks — they already do. But automation isn't the same as adoption and volume isn't necessarily value.
The question isn't whether AI can ape into tokens faster than humans. We know they can.
Rather, our focus should be on whether all this artificial activity brings us closer to financial sovereignty and adoption.
Here’s a spoiler alert: It doesn’t.
And, at the end of the day, if the tools we create don’t help us achieve our ultimate goal, why use them?
Well, the article has an answer for that as well. And that’s where most of my concerns lie.
The Hidden Threat of Centralization
The existential question of AI replacing humans in the crypto market is just a distraction.
Rather, the real concern is this: AI agents, while more autonomous, are still wielded by people. And so far, the majority of AI capability has been centralized in the hands of major centralized authorities — the very same ones that crypto was meant to cut out.
And that will be the core of the real question: Will AI agents enhance individual autonomy … or become yet another mechanism of institutional control?
See, until now, AI agents have been seen as a force for decentralization.
That’s because they allow us to tap into the AI boom … without being beholden to the few tech giants and governments that can afford AI development.
As Haseeb Qureshi, managing partner at Dragonfly Capital, asserts in the Decrypt article: "Instead of having to raise millions of dollars," developers could "launch an application with $10,000 of AI cloud compute."
And that’s exactly why AI crypto projects have exploded in recent months.
But there are two main problems with this.
First, these AI agents need to actually compete with the AI programs developed by those centralized power houses.
As tired as we may be watching AI giants like Microsoft and Google dominate every emerging technology, they are the ones able to push the envelope and develop the next best model thanks to their resources.
Right now, most AI agents in crypto aren’t doing anything that the AI models run by tech giants can’t do. And the sad reality is that people typically choose convenience over principles. We can wave the decentralization flag all we want, but if OpenAI's latest model runs circles around the decentralized, open-source alternatives, I know where users will flock.
Here’s a hint: It won’t be the decentralized models.
That means we must improve the use and function of decentralized AI projects beyond the standard capabilities we currently have.
Second is a concern Qureshi brings up in the article itself. He goes on to predict a more concerning future where "chatbots will start hiding that they are AIs" and "pass as humans."
The threat of not knowing which projects are human run and which have AI agents at the helm will likely spark calls for verification.
And that’s where things get tricky: While protecting against AI manipulation, we risk recreating the very systems we sought to escape.
See, crypto was intended from the beginning to be trustless. You don’t have to reveal your identity — for the most part — because of that trustlessness. This way, no centralized authority can intercede on your financial freedom.
This is in stark contrast to the way the TradFi financial system operates.
Think back a few years, and you’ll be able to recall what that looks like. During the COVID lockdowns in Canada and the sanctions against Russian leaders at the onset of the conflict in the East is exactly what it looks like when a centralized authority seizes your money.
Every identity verification layer, no matter how decentralized, introduces new gatekeepers and the need to trust the platforms you interact with.
It will require a delicate balance between maintaining system integrity and preserving crypto's permissionless nature. It’s a philosophical and technical challenge. And it’s one we will be asked to face in the near future.
So where does this leave us?
The fusion of AI and crypto isn't inherently flawed. In fact, I fully stand behind my earlier stance that AI will push crypto into its future.
However, how and when we use AI in crypto is critical.
Right now, we’re watching history repeat with this groundbreaking technology. The same centralization patterns that plagued both web2 and early crypto have begun to creep into this new AI-crypto landscape. This time, it is hiding under a banner of “democratization” … all while using centralized infrastructure.
The answer isn't another whitepaper or tokenomics model. It's about building systems that preserve human agency while leveraging AI capabilities and true democratization.
I know there are a lot of crypto AI projects these days. And I expect more as this bull run continues.
So, I hope when the next time you research a new crypto AI to invest in, you keep in mind these roadblocks and goals for the sector. And be sure to check out my colleague Dr Bruce Ng’s tools for how to narrow down your crypto AI options.
Best,
Jurica Dujmovic
P.S. While the battle for AI is still being waged, one area where crypto outperforms TradFi institutions is yield.
Money market rates average 0.6% ... CD rates are at 1.75% and falling fast ... and S&P yields are at a two-decade low of 1.18%.
But with her crypto strategy, DeFi expert Marija Matić has been able to target yields as high as 101%, 912% and 1,168%.
And this coming Tuesday, Jan. 14 at 2 p.m. Eastern, Marija will host her Superyield Income Summit where she’ll break down exactly how she does it.
It’s completely free to attend. Just be sure to save your seat now.
Tidak ada komentar:
Posting Komentar