Hook
On a quiet Tuesday in Zagreb, an autonomous Uber vehicle logged its first commercial trip. No press release named the technology partner. No technical specifications were published. No safety report was filed with the public. The only thing we know is that Uber has launched a robotaxi service in the Croatian capital—and that the entire event was reported with the same breathless enthusiasm that greets a new DeFi token launch.
But here is the paradox: the very technology that enables this milestone—autonomous driving—is a perfect case study in centralized control. The vehicle is a black box. The algorithm is proprietary. The data flows to a single corporate server. The decision-making is opaque. And yet, we in the blockchain community often celebrate the arrival of autonomous vehicles as a step toward a decentralized future. We are wrong.
“We built the temple, but forgot who the god is.” The god here is not the user, not the community, not the open-source protocol. It is Uber. And the temple is the vehicle itself, a closed system that reminds us more of Web2 than Web3.
Context
To understand the Zagreb deployment, we must first understand Uber’s relationship with autonomy. In 2020, Uber sold its Advanced Technologies Group (ATG) to Aurora Innovation, a move that signaled a strategic pivot from building its own self-driving technology to becoming a platform that integrates third-party solutions. This is a classic “asset-light” strategy: Uber provides the demand network, the insurance, the regulatory navigation, and the user interface, while a technology partner provides the autonomous driving stack. In Las Vegas, Uber partners with Motional. In San Francisco, it partners with Waymo. In Zagreb, the partner remains unnamed—a detail that speaks volumes.
The lack of transparency is not accidental. Uber’s platform model allows it to switch partners, negotiate better terms, and avoid the capital expenditure of owning a fleet. But it also means that the technology behind the rides is a black box. The user trusts Uber, but Uber itself trusts a third-party vendor whose code is not open, whose training data is not shared, and whose safety metrics are not independently audited. This is the antithesis of the decentralized ethos that blockchain champions.
In the blockchain world, we talk about “code is law” and “trustless verification.” A smart contract on Ethereum is transparent; anyone can audit it. The autonomous vehicle code in Zagreb is a corporate secret. The difference is not merely technical—it is philosophical.
Core
Let me break down the Zagreb deployment through the lens of my own experience as an open-source evangelist who has spent years analyzing the intersection of code and values. I have audited tokenomics, studied DAO governance, and written about the ethical implications of decentralized finance. The Uber case is a mirror that reflects the blind spots of our own industry.
Technical Transparency vs. Proprietary Black Boxes
The first dimension is technical. The analysis of the original article (the parsed content I received) revealed that the Uber deployment contains zero technical details: no sensor configuration, no HD map dependency, no mention of safety drivers, no partner identity. This is not a minor omission—it is a feature of centralized control. In a decentralized system, the code is the contract. In Uber’s system, the contract is hidden.
I recall a conversation with a colleague who worked on the open-source autonomous driving stack from Baidu’s Apollo project. He told me that the biggest challenge was not the technology, but the culture of secrecy. “We have to fight the instinct to hide our failures,” he said. “In open source, we share everything—the crashes, the edge cases, the training data. In proprietary development, you only share the wins.” This asymmetry is dangerous. When the code is closed, the user cannot verify safety. The regulator cannot audit the algorithm. The community cannot contribute improvements.
In Zagreb, the lack of technical details suggests that the partner is likely a startup with an immature stack, or one that operates under a non-disclosure agreement that prevents them from revealing their involvement. Either way, the user is left in the dark. This is not the transparency that blockchain promises.
Ethical Liability and the “Code is Law” Fallacy
The second dimension is ethics. The parsed content correctly identified that the biggest risk is a safety incident. Uber’s history—the 2018 fatal crash in Tempe, Arizona—looms large. But the ethical question is not just about safety; it is about accountability. In a decentralized autonomous vehicle network, if a smart contract causes an accident, the code is the culprit, and the protocol’s insurance pool covers the loss. In Uber’s model, the liability is corporate. But who is responsible? The technology partner? Uber? The driver (if any)? The answer is unclear, and that ambiguity is a feature of centralized power, not a bug.
I think about the blockchain principle of “code is law.” It is a powerful idea: the rules are encoded, immutable, and apply equally to all participants. But the Zagreb autonomous vehicle is not governed by code; it is governed by a corporate contract. The user cannot inspect the code that decides when to brake or turn. They cannot fork the code if they disagree with the decision-making. They are subjects, not participants.
Data Ownership and the Surveillance Economy
The third dimension is data. Every autonomous vehicle generates terabytes of data per day: location, behavior, biometrics, road conditions, interactions with pedestrians. In a decentralized system, that data could be owned by the user, stored on a distributed network, and monetized through a tokenized marketplace. In Uber’s system, the data flows to a central server, where it is used to train the algorithm, improve the product, and—eventually—sell insights to third parties. The user gets a ride; Uber gets the data. This is the same extractive model that Web2 perfected.
I have written before about the “data prison” of centralized platforms. The user contributes value but receives no ownership. The platform accumulates power. The Zagreb deployment is a perfect example: the user does not even know who is collecting their data, because the partner is anonymous.
Contrarian
Now, let me play the contrarian. The blockchain community often romanticizes decentralization as the only path to a just future. But the reality is that centralized systems can be more efficient, safer, and more accountable in the short term. The autonomous vehicle stack is incredibly complex. It requires massive amounts of training data, expensive hardware, and rigorous safety testing. The open-source alternatives—like Apollo or Autoware—are still far from commercial readiness. It is possible that the proprietary, centralized approach is the only way to achieve the safety and reliability required for public deployment.
Moreover, the platform model allows Uber to rapidly deploy in multiple cities with different partners, adapting to local regulations and conditions. A decentralized autonomous vehicle network would require a global consensus protocol, a tokenized incentive system, and a governance mechanism that could handle the complexity of driving in diverse environments. We are not there yet.
But the contrarian view misses a crucial point: the path we choose now shapes the future. If we accept the centralized black box as the norm, we will never build the decentralized alternative. The infrastructure, the data, the regulatory frameworks—all will be optimized for the centralized model. The blockchain community must recognize that autonomous driving is a battlefield for the future of mobility, and if we do not fight for openness and transparency, we will lose the opportunity to encode our values into the code.
“Authenticity is a signal lost in the noise.” The authenticity of a decentralized system is its openness. The noise of Uber’s marketing obscures the fact that the user has no control.
Takeaway
So what is the takeaway from the Zagreb launch? It is not that Uber is evil, or that autonomous driving is bad. It is that we must hold the blockchain community to a higher standard. We must demand transparency from the technologies we use, even when they are not built on our protocols. We must question the black boxes, whether they are proprietary or corporate. And we must build the alternatives.
“The ledger remembers, but the heart forgets.” The ledger of blockchain technology remembers the promise of decentralization. But our hearts are easily seduced by the convenience of a centralized ride. We must not forget the values that led us to this space.
I propose a concrete action: the blockchain community should fund an open-source autonomous vehicle stack that is fully transparent, auditable, and community-governed. It may not be ready for commercial deployment in Zagreb tomorrow, but it will be a beacon of what is possible. We did it with DeFi, we did it with NFTs, we can do it with mobility.
The future of autonomous driving is not yet written. The code is not yet law. But the choices we make today—the partners we celebrate, the questions we ask, the transparency we demand—will determine whether the temple serves the people or the god.
“Faith in the protocol is not faith in the people.” Faith in a protocol is faith in the code. Faith in the people is faith in the values they encode. In Zagreb, the code is hidden. The people are passengers. The values are corporate. We must do better.