I participated in @Mira_Network a few months ago, and recently saw that they launched a Kaito Yaps campaign, and I have also swiped a lot of related tweets, so I would like to write about my reasons for choosing to invest in Mira. Project Introduction To understand what Mira is for, it is first necessary to understand the illusion problem of AI models, which is that AI produces completely fabricated information that leads to inaccurate and untrue results. This phenomenon can be caused by defective training data, lack of proper basis for the inference process, bias in the model itself, etc. Correspondingly, researchers have proposed a variety of methods to mitigate the illusion problem and improve the reliability of AI output, such as introducing human review, rule-based filters, self-validation, and ensemble models, but these methods also have their own limitations. In this regard, Mira combines the consensus mechanism of the blockchain and proposes a solution: each AI output is broken down into smaller factual claims, which are independently evaluated by multiple validators in the Mira network, each node runs its own AI model, votes to determine the authenticity of the claim, and finally decides to agree, flag, or reject the claim based on whether there is a disagreement based on the judgment of multiple models. In this process, there is an on-chain traceable record for each verified output, so Mira's core function is a decentralized audit/trust layer. If you want to learn more about Mira's role in AI validation, you can read this @MessariCrypto research report: Team & Financing Background Founder @karansirdesai and Chief Growth Officer @stone_gettings are both from the well-known venture capital firm @Accel, the VC's most well-known investment is Facebook, and with this relationship, Mira has also become Accel's first Web3 project in three years. Karan spent nearly three years at Accel as an Investor and prior to that as an advisor at The Boston Consulting Group (BCG): Ninad Naik, Chief Product Officer, was previously a product director at Uber and spent nearly 10 years at Amazon AI, leading collaborative organizations in terms of product and management: In terms of funding, Mira received $9 million in seed funding from Framework, Mechanism Capital, Bitkraft and others. In addition, Mira has been selected as one of the top 100 AI startups in 2025 by CB Insights@CBinsights, an AI data services company: Partners Among the projects that have reached a partnership with Mira, the following projects have been logged into Binance: - @kernel_dao: Integrate Mira's consensus mechanism into the BNB ecosystem. - @PhalaNetwork: Mira is the official model provider for Phala Network's AI agent, providing verifiable LLMs and trustless inference. - @lagrangedev: Lagrange's zkML prover supports Mira's verification layer. - @plumenetwork: Mira's verification infrastructure will be integrated with Plume's AI framework. In addition, Mira's node operators include: Project Progress In the data released by the project team in May, Mira has achieved the following results: - Increase the accuracy of AI answers from 70% to 96% through consensus verification - Handles an average of 5B token verification requests per day - The overall ecosystem covers more than 4.5 million users In June, there were also the following developments: - Partnered with Delphi to improve the verification capabilities of AI-generated content - Launch the Kaito leaderboard - Partnered with GoPlus At present, the methods of participating in Mira include interacting with the AI chatbot Klok, participating in Kaito yaps, etc., and continue to look forward to Mira's next progress.