Bankless Times talked to Matthijs de Vries, the founder of Nuklai, about how AI and Web3 are changing the future.
Can you tell us a bit about Nuklai and what inspired you to create it?
Businesses are sitting on mountains of data, but current tools make it difficult to access and use that data effectively, especially for AI development. Existing data marketplaces are fragmented and lack interoperability. Nuklai is here to bridge that gap.
We’ve built Nuklai as the world’s first on-chain smart data ecosystem. Think of it as a secure platform where businesses can share and access data relevant to their needs. But it's not just about sharing raw data. We focus on helping businesses turn their data "smart data" – enhancing the quality of data to unlock its true value, and using that data to train AI models. This will lead to internal innovation, monetization, and more ways to profit through artificial intelligence.
How does the NAI token work in tandem with Nuklai’s infrastructure?
As a Layer one project focusing on data collaboration, data sharing, data enrichment, data control, and AI training, we replace the traditional middle-person (centralized entities) with a decentralized infrastructure. This has its advantages. For example, the ‘power’ over the network is distributed, meaning every participant owns it. There are also a good number of efficiency benefits to decentralization, as many overhead costs are automated.
At the same time, decentralization of infrastructure means that those who participate in it, such as running Nuklai nodes, must be rewarded for that work. This is where the NAI token comes in, as gas fees are paid in NAI. Early node-runners will receive additional NAI to bootstrap the network, so it can be very interesting for individuals, groups, and businesses to run nodes.
It is also good to know that we have a unique infrastructure where NAI plays a role. First, there is an Orchestration layer, which receives the work and finds the right clusters to execute the work. Once the work is done, the cluster validates it and notifies the network. There also is an Execution layer, which consists of many different clusters (such as GPU partners and storage). This is where the work is done, and logically NAI also plays a role here.
For enterprise use, private and/or semi-private subnets can be created. These are directly connected to the mainnet; NAI is needed to develop these subnets. Subnets are super interesting for enterprise use cases as they enable enterprises to collaborate, while owning their subnet ‘together’ and solving part of the trust problems that generally occur when companies need to work together on data.
Nuklai’s successful launch is quite an achievement. What do you think drew so much attention to the project?
It was a perfect storm for the Nuklai, really. Our various partnerships, such as with Chainlink, Ava Labs, and Omniscia, helped bring attention to the token launch. We also partnered with enterprise businesses like Crunchbase and CubeMatch, and have a beta version of our product already available. Our beta is showing great potential, as our first Web3 marketing smart dataset through Bitcoin Research boasts hundreds of contributions from the community. BeInCrypto featured Nuklai as the top project in its viral article, which certainly didn’t hurt.
We also got selected to join an exclusive startup accelerator Startup Autobahn where we rubbed shoulders with industry giants including Porsche, Toyota, and Bosch. We feel blessed to say that our project is generating a lot of excitement – and we’re only just getting started!
Can you elaborate on what exactly “smart data” entails and why it's a game-changer for AI development? And how does Nuklai help transform raw data into smart data?
Smart data is like highly refined fuel for a fast car. A high-end vehicle can’t run on regular fuel for long, just as a powerful AI can’t work with regular data. For AI models to be effective, they need smart data. An AI with smart data can generate new insights that lead to innovation. Companies can also monetize an AI’s findings to create additional revenue streams.
It all boils down to context. Imagine a spreadsheet – the column names are like basic instructions for the data, but this isn’t enough info for AI to be effective. Nuklai lets users add more details, such as data descriptions, limitations, and flags for when data quality might change. This extra info, called "metadata," makes the data "smart." Now, AI models can understand the data without needing as much hand-holding. Plus, metadata helps automate data labeling processes, which is a big step for AI training.
The best part is that we can tailor it to different needs. There's clear, human-readable stuff for data scientists or specific instructions for large language models. Collaboration is critical, too, so groups working together on projects can optimize based on metadata results.
What is the future of AI in data decentralization?
AI is taking center stage and will likely be the driving force behind data systems in the coming decade. Decentralized solutions are also on the rise: we’re already living in a future where Web3 players and innovators integrate AI into their decentralized projects.
At Nuklai, we see a massive opportunity for decentralized data solutions, particularly in industries like IoT and automotive. In the future, Nuklai will serve as a go-to infrastructure for businesses. What we are developing is basically an App Store, but for data. With our execution layer, our clients will be able to enhance their data through handy features like Al, distribute computing, anonymization and so on. Our Layer 1 blockchain will act as a non-intrusive layer on top of existing infrastructures: organizations will just need to grant Nuklai's decentralized network access to their raw data.
Does this future entail any risks?
As with any company developing in the field of AI and decentralized data – especially when attempting meaningful work – there are always compliance, security, privacy, and other risks. Data breaches could be devastating if hackers gain access to these systems. Also, concerns about surveillance and unforeseen ethical risks may rise as these technologies evolve.
Those risks fuel our dedication to releasing high-quality products. Our infrastructure is built to mitigate threats and will contain malicious activity in a small cluster, preventing its spread throughout the ecosystem. We also make sure never to store our users' private data. In the unlikely event of an attack, our users' risk is minimal at most.
As we address these challenges, thousands of businesses can unlock valuable insights from their data, leading to groundbreaking innovations. Think about the endless potential for personalized healthcare or the creation of entirely new industries. With open communication, collaboration, and a commitment to responsible development, we can navigate the potential risks and harness the incredible potential of this future.
Contributors
