Exclusive interview | xTAO founder said Bittensor hopes to become Bitcoin in the field of artificial intelligence

According to Bijie.com, Karia Samaroo, founder and CEO of xTAO, explained why AI needs to be decentralized. xTAO is the only listed company dedicated to the Bittensor ecosystem. Artificial intelligence has attracted the public's imagination at an unprecedented rate. However, under its potential, serious concerns about concentration of power and control are hidden. Currently, the most popular AI models are the exclusive property of several large tech companies that completely control the design and use of these models. Crypto.news interviewed Karia Samaroo, founder and CEO of xTAO, a publicly traded company dedicated to the decentralized AI ecosystem of Bittensor (TAO). Samaroo explains why an alternative model is needed to make AI more open, decentralized, and meet users’ needs. crypto.news: What does blockchain bring to AI, and what role does Bittensor play? Karia Samaroo: Centralization is the biggest problem with AI. As AI develops into the most powerful tool ever created by humans, only a few companies control it creates huge concentrated risks. I often liken Bittensor to Bitcoin. Bitcoin solves the centralization problem around currency: it won’t be inflated, it can be accessed by anyone, and there is no gatekeeper. Bittensor applies the same idea to AI. For centralized AI, such as OpenAI, an institution decides how to train a model, what data to use, what biases are there, and what to review. They can also cut off access at any time. This is a big problem. Bittensor uses Bitcoin's model to solve this problem of AI. CN: How do companies introduce decentralization into AI? KS: There are some good examples of decentralized AI solutions. Grass motivates data collection, although it focuses on a part of the AI ​​stack. Render is a decentralized computing network, which is also very important. Bittensor has a wider range. I would call it “the global network of AI.” It focuses more than just one area like data or computing. It has multiple subnetworks, each solving different problems in the AI ​​stack, and they are all connected to each other. CN: Why do companies build on Bittensor instead of using more mature models like OpenAI? KS: I think there are several reasons. One is philosophical. Many people built on Bittensor see the value of contributing to the decentralized networking and decentralized AI mission. This definitely has a lot of appeal. The other is technical. In decentralized networks, scalability has advantages. Bitcoin, for example, created the world's largest computer through its incentive mechanism. It is so wide that it will never be turned off because it has many nodes running on different locations, different networks and power supplies. Then there is the concept of open innovation. Anyone can experiment, iterate and monetize their models without the need for a gatekeeper. If you are an AI engineer, you usually have to apply for a job, attend an interview, be hired, and then eventually work on a very specific task within the company. On Bittensor, you can choose a subnet on which you want to mine, build your model, compete with others, and get paid immediately. CN: AI models run by large tech companies benefit from a lot of data, such as Grok owns Twitter. How does decentralized AI compete? KS: I think Grass is a good example, and there are similar projects on Bittensor. The idea is to crowdsource data and inspire people to collect and manage data. The network has evolved very significantly. This is how decentralized networks introduce equal or even higher quality data sets. Large tech companies control the richest data today, but with the right incentives, decentralized systems can compete with it. Another big problem is that when Meta or Twitter has your data, you get nothing back. As a contributor, you won't get rewards. Decentralized networks subvert this. They align incentives with creators and contributors, and that’s how they should be. If you take a photo, you should be signed. If you published an article, you should benefit from it. CN: How does decentralized AI solve the security and social impact issues of its models? KS: There are several aspects to safety. One is training data. If it is biased, toxic, or contains sensitive information, then there is a problem, which is true for both centralized and decentralized systems. This is a problem that people are working hard to solve every day. The other is the output of the model. How do you prevent harmful output? In Bittensor, this is handled by the validator. They are responsible for detecting harmful or low-quality outputs, and the better they do, the more rewards they receive. It has been incorporated into web design. The foundation also has some monitoring policies, but the goal is to phase out these policies. Over time, security and governance have indeed become the work of validators. CN: Are you worried that these models will be under scrutiny in the future, whether from government or as a response to biased output? KS: That's a good question. I would compare it to centralized or state-owned media where a single decision maker has the option to show or not show what. If they are under pressure or just make decisions internally, they can change the way the output looks. This is a big problem. We've seen this on social media already. If Meta wants to push for some narrative, they will do it. It's not necessarily evil—it's just how incentives work. Decentralized AI can better represent the people. It’s not perfect, but if a subnet or product on Bittensor becomes too biased, participants in the network can vote and adjust incentives. This means that underperforming will receive fewer rewards. The idea is that if the system reflects the population, people will support products that feel fair and transparent. And it's easier to audit – you can see the incentive structure, you can see the code. With closed systems, you can't do this. This is why people are worried about centralized AI.