While AI and Blockchain converged, one project stands at the intersection of this revolution: OpenLedger. With promises to give the decentralized character to train models, assign data, and energy agent economies, OpenLedger is moving towards a new era of transparent and owned artificial intelligence.
In this interview, we sat with the main contributor to understand the principles, innovations and the road map that supports the unique OpenLedger.
Ishan Pandness: Hi Kamesh, it is a pleasure to welcome you in the “Behind the Startup” series. Can you start telling us about yourself and how you participated in OpenLedger?
Camsh: Hi Ishhan, I am a key shareholder in OpenLedger. Before OpenLedger, I was working on an Amnesty International’s research and development company, during which I worked with institution agents such as Walmart, Cadbury and more. There were a few of the main problems that we noticed in the artificial intelligence clip, which are black box models and a lack of transparency, and that there was no way to know the data that led to a specific conclusion. The biggest issue is that the central companies have not somewhat compensated for the contributors to the data, and this is what we are trying to process with OpenLedger.
Ishan Pandness: OpenLedger is placed as “Blockchain designed for artificial intelligence”. Can you walk through the gap in the current infrastructure that OpenLedger tries to solve and why this moment in time?
Camsh: As we mentioned earlier, central companies trained artificial intelligence models on user data without permissions and made billions of dollars without paying anyone fairly. On OpenLedger, you leave each training step and a source of data and upgrading the form that anyone can examine. This is now important because people ask artificial intelligence financial advice, health suggestions and even electoral coverage. While dealing with such sensitive topics, it is important to ensure that the model uses precise data, not hallucinations. Using attribution proof, we can eliminate data that led to a specific harmful conclusion, and ensure safety in sensitive use.
Ishan Pandness: One of the most ambitious ideas behind OpenLedger is the creation of data -based bonuses and typical contributions. In practical terms, how to measure the contribution to a decentralized environment?
Camsh: Think of a common history on the series, each group records data and a model with the wallet it presented. When the network trains a new version or responds to the user’s inquiry, it looks through this date to know the contributions that were involved. Every time your entry appears, you automatically receive part of the fees associated with this procedure. This information is general, so that anyone can open the explorer and follow exactly how to use their work and what it got.
Ishan Pandness: Let’s talk about the model and OpenLora factory. From a technical perspective, how are these tools designed to deal with resource sharing, GPU bottlenecks, and extensive extensive repetition demands?
Camsh: Think about Model Factory as a platform not a symbol where anyone can set a specialized language model without renting a full data center. You can choose a basic model and identify parameters. When you’re done well, it is saved as a lightweight Lora transformer, where many versions can live side by side without eating huge amounts of memory or frequency range. Then Openora allows you to connect these transformers to a common basic model during reasoning, so that one GPU can switch between dozens of disciplines, allowing frequent repetitions. The model and OpenLora factory are very important pillars for the ecosystem because it allows everyone to participate in developing artificial intelligence without any great cost.
Ishan Pandness: You also offer a concept called “POA)”, what exactly is measured here, and how do you make sure that it is a trusted measure in assessing the activity of the worker?
Camsh: Proof of support is how to track the data used by the model to access a specific conclusion and reward each meaningful contribution. When using your data with a form to create an inference, it is recorded on the chain. Each time users depend on the form, part of the revenues are directed automatically, and the entire track is opened for anyone to verify it. The shareholders are allowed to see evidence of their work on the chain and allow them to reward it fairly.
Ishan Pandness: Amnesty International Kings as a concept depends on long -term tracking and confidence. How do you plan to deal with form problems, use the agent, and assign the value of the estuary through artificial intelligence agents?
Camsh: Our priority is to ensure that shareholders are always compensated fairly. To avoid problems such as using models and using the agent, we will host all models ourselves and external access will be across API only.
Ishan Pandness: I glimpsed to Testnet. Can you share the following group of landmarks that the milestones are working on, and what developers can expect when engaging with OpenLedger in its current stage?
Camsh: The shareholders already rotate the contract and flow of real data in the network. We currently have more than 4 million active knots that work on our Testnet, and we have just finished from EPOCH 2. We have more than 10 projects based on us, including a former Google DeepMind. We are very excited to share that we will soon go to launch TGE and Mainnet. We will share the full details soon, so stay away.
Ishan Pandness: Finally, as a person who builds infrastructure for information technology, what advice you give to developers or researchers looking to enter the decentralized AI intersection?
Camsh: The best advice I can give to people is to keep your product simple. People should be able to understand what you do in the first few minutes. Users are more interested in whether the product works smoothly and solves their problem instead of luxury keywords.
Do not forget to love and share the story!
Detection of the acquired interest: This author is the publication of an independent shareholder across Commercial blogging program. Hackernoon reviewed the report on quality, but the claims here belong to the author. #Dyo
