Nesa, a Binance MVB Incubated Layer 1 Blockchain Platform, is Poised to Democratise AI and Open it Up to the World

Nesa Mascot Creation Contest on the Lookout for Winning Design

Nesa, a Layer 1 blockchain incubated by Binance's Most Valuable Builder (MVB) programme and the first to integrate over 1,000 AI models on-chain, has recently unveiled the Nesa Mascot Creation Contest in collaboration with Stability World AI, a generative artificial intelligence (AI) platform designed for Web3.

Nesa is officially announcing over 1,000 AI models on Day 1 launch. Every model supported on Nesa is sequentially sharded with BSNS technology for consumption on low-powered devices – a feat never before seen at scale.

Welcome to the Global Repository for AI. It's time for the… pic.twitter.com/MlxCMQruls
— Nesa (@nesaorg) May 1, 2024

This contest opens up an exciting opportunity for individuals to craft a mascot that encapsulates the essence and values of Nesa, utilising AI to enhance community involvement and spark design ingenuity within the Nesa ecosystem.

The contest breaks down barriers to entry, ensuring that artistic expertise is not a prerequisite.

Today we are kickstarting the Nesa Mascot Challenge together with @StabilityW_AI! Show off your creativity and shape Nesa's brand story!

Join us now! Win a Nesa Playground Whitelist spot and more!

📝Official rules and registration on Nesa Discord https://t.co/YjVFNGOFQb
📅… pic.twitter.com/8ttC0M8ax3
— Nesa (@nesaorg) August 6, 2024

Stability World AI' text-to-image technology offers an accessible tool for participants to translate their creative concepts into captivating mascot designs.

This approach democratises the creative process, enabling a diverse array of talents to contribute to the evolution of Nesa's brand identity.

Jay, BD Manager at Stability World AI, expressed his excitement:

“We're thrilled to partner with Nesa on this innovative competition. By providing our advanced AI technology, we aim to empower creators and foster a vibrant community experience within the Web3 space.”

🔥 Stability World AI's collaboration in powering the @nesaorg Mascot Creation Contest has been featured on @cryptodailyuk! 📰

🔗 Discover how we're driving innovation in Web3: https://t.co/SRSsHYrsX8#StabilityWorldAI#GenerativeAI#GentoEarn
— Stability World AI (@StabilityW_AI) August 8, 2024

The Nesa Mascot Creation Contest concludes on 17 August, this Saturday, with winners determined by a blend of community voting and the verdicts of a judging panel.

The victorious design will be adopted as Nesa’s official mascot, and the top five entries will be granted exclusive whitelist access to the Nesa playground.

Moreover, every participant is rewarded with 100 Stability World AI credits and an additional 1,000 Nesa points for submitting their designs on Discord, ensuring that all creators are recognised for their contributions.

Nesa, a Binance MVB-incubated Layer 1 blockchain

On 1 March, Binance Labs, the venture capital and accelerator arm of Binance, unveiled the 13 early-stage projects chosen for the seventh iteration of the Most Valuable Builder (MVB) Accelerator Programme, among which was Nesa.

This 10-week accelerator programme is a collaborative effort between BNB Chain, the community-driven blockchain ecosystem encompassing the world's largest smart contract blockchain, and Binance Labs, with the addition of CMC Labs as a strategic partner for this season.

Season 7 of the MVB Accelerator Programme featured two distinct tracks: the MVB Startup Track and the MVB Founder Track, with the startup track commencing officially on 5 March.

DeFi: @bitu_protocol, @blumcrypto, @surf_protocol, @vooi_io

Infra: @Aggregata_xyz, @nesaorg, @Nimble_Network

Application Layer: @aspecta_id, @HoloworldAI, @opinionlabsxyz, @SideQuestWeb3, @StoryChain_ai, @tiltedstore
— Binance Labs Fund (@BinanceLabs) March 1, 2024

The programme is designed to facilitate the growth of early-stage Web3 projects within the BNB Chain ecosystem, culminating in potential investment from Binance Labs based on Demo Day presentations and the overall performance of the projects throughout the MVB program.

Yi He, Binance Co-Founder and Head of Binance Labs said:

“Binance Labs remains committed to supporting early-stage founders. As part of this commitment, MVB is dedicated to bolstering and advancing projects within BNB Chain to grow the Web3 ecosystem.”

The MVB Accelerator Programme remains a highly competitive endeavour, with Season 7 attracting over 700 applicants.

Only 1.8% of the total applicant pool was selected, leading to the admission of 13 teams into the MVB Season 7 Accelerator cohort.

Following the completion of the 10-week accelerator programme, the 13 project teams had the opportunity to showcase and pitch their projects to investors at Demo Day.

Join us to witness innovation at the Demo Day of the Most Valuable Builder (MVB) VII.

⏰ 2-3PM UTC
📅 Tuesday, April 16
📍https://t.co/FaHixMhQ4T

More details 👇https://t.co/54zKws1r3z
— BNB Chain (@BNBCHAIN) April 9, 2024

You can view Nesa's demonstration from 32:03 to 38:29 in the YouTube video below.

Nesa: Pioneers of Decentralised AI Inference

In 2024, the convergence of AI and Web3 has emerged as a pivotal theme in industry advancement.

Nesa, positioned as a lightweight AI Layer1, is at the forefront of redefining privacy and security standards for decentralised AI inference.

By streamlining inference steps within an encryption-optimised framework and harnessing cutting-edge technologies such as Zero-Knowledge Machine Learning (ZKML) and split learning, Nesa effectively tackles the privacy and security challenges inherent in centralised AI platforms.

We believe that equal access to powerful technology is a human being’s right. And we think that the solution to the problems mentioned will be achieved by decentralizing inference and merging crypto with AI. It is at this intersection that the purest forms of ownership, freedom,…
— Nesa (@nesaorg) April 19, 2024

Nesa, whose name signifies a "miracle," employs the Artificial Intelligence Virtual Machine (AIVM) framework to seamlessly integrate on-chain assets with off-chain activities in a unified execution environment.

The platform's mining network achieves consensus and disseminates reliable proofs on-chain through system oracles, thereby guaranteeing the integrity of AI inference outcomes.

Two innovations set Nesa apart: the AIVM and its privacy-centric decentralised inference protocol, underpinned by ZKML.

The AIVM is a comprehensive, decentralised system crafted to meet the demands of real-world AI applications, while ZKML ensures transaction confidentiality and verifiability by implementing bifurcated transactions at the blockchain level.

At Nesa's core is its decentralised query marketplace, a trailblazing initiative that introduces a GPTs store for Web3.

This marketplace, driven by a reward economy, motivates AI developers, queriers, miners, and model reviewers to contribute to a global, decentralised, and immutable repository of AI models on-chain.

Nesa’s solution is the first 100% private critical inference, encrypted end-to-end, and distributed across a network of low level machines powering the most demanding models in the world. This is an achievement never before done at scale, and it has been made possible through…
— Nesa (@nesaorg) April 19, 2024

Nesa's mission is to transcend the limitations of centralised AI platforms by offering an open, secure, and efficient AI ecosystem.

Harry Yang, Co-Founder & CTO, analysed:

“Nesa places a high emphasis on the technical requirements for security and privacy, ensuring all added operational requests enhance project privacy and effectiveness.”

It achieves this by committing computation inference results to the blockchain through a security committee.

These results are then validated by consensus after being summarised and voted on through smart contracts, with nodes being rewarded or penalised accordingly, ultimately delivering consensus results to users.

Yang explained:

“In addition, Nesa is currently developing new technical products such as Speed Learning and Watermarking Of The Models to further enhance the privacy and trustworthiness of the technical solutions combining software and hardware.”

To attract more developers to the Nesa ecosystem, the team has introduced NES tokens, leveraging the platform's influence and operational efficiency to minimise users' gas fees and ensure a safe and private usage experience.

Its testnet is currently open to the public, registering 224,000 active addresses and over two million operations a week after its launch.

Today is the official launch of Nesa Testnet.

Due to high demand, we are inviting whitelisted registrants in phases over the next two weeks.

1) Each day, check your email for an invitation.
2) When you receive your invite, register your account on https://t.co/bz6TbJTtKi… pic.twitter.com/cTIp2YVrRy
— Nesa (@nesaorg) May 18, 2024

Yang pointed out:

“Advanced AI intelligent applications such as Sora models and video generation are being gradually developed and tested, supporting the development and optimization of decentralized AI inference. At the same time, we will seriously listen to the voices of community developers and continuously optimize the project.”

Currently, Nesa is part of Google Cloud's AI layer plan, receiving support for servers and computing resources to collectively shape the future of decentralised AI development.

Today, Nesa was admitted into the Google Cloud Program AI Tier and given a grant for servers and compute. We're happy to be receiving support from Google to build a decentralized AI future. ⚡️ pic.twitter.com/ZP6fwqciRw
— Nesa (@nesaorg) May 16, 2024

Nesa's Story

Nesa was conceived when a group of friends encountered frustration while using a top-tier centralised Large Language Model (LLM) for their project queries.

Despite its industry-leading status, the LLM's performance was inconsistent, varying daily for identical prompts and showing signs of regression over time.

Lacking transparency into the reasons behind these fluctuations, and with no recourse for dialogue or control over the centralized model, they felt powerless and at the mercy of the black box system.

This unpredictability was detrimental to their research, time, and business operations.

In response, they assembled a world-class engineering team to develop the first decentralised, open, and trustless AI network, addressing the issues they faced.

The team comprises elite members from diverse fields, including university professors, technical engineers, and scientific researchers, all of whom are adept at contributing to privacy inference and protection technologies rooted in AI principles.

Nesa's mission is to bring AI onto the blockchain for the benefit of our future.

With AI on-chain, its performance can be transparently tracked, verified, and shared by all, and its outputs can be readily monitored, evaluated, and controlled for the first time.

As AI becomes increasingly pervasive in our lives, individuals, businesses, and society at large require an open and decentralised AI framework.

Nesa achieves this by seamlessly integrating privacy, security, verification, and consensus through a proprietary end-to-end system that is both easy to integrate and remarkably fast for querying.

Nesa represents the world's solution to the growing ubiquity of AI in our daily lives.

The team stated:

“Nesa means miracle. Our name is a tribute to the golden age of AI that we are living in. The logo of NES token is symbolic of the miracles that trusted AI will deliver for humanity, and a reminder of how advanced technology is truly indistinguishable to magic.”

Harry Yang, Co-Founder & CTO

Yang, formerly with Facebook AI, led the development of Make-a-Video, Meta's groundbreaking tool that enables users to instantly create videos from text using AI.

With deep expertise in LLM design and computer vision from his time at Meta, Yang's work has earned thousands of citations and numerous highlights in IEEE and CVPR conferences.

His outstanding achievements have also earned him an invitation for an AI professorship.

Yang holds a PhD in AI from University of Southern California (USC) and a Master of Science in Computer Science.

His visionary contributions exemplify the innovation driving Nesa's mission to transform the field of AI.

His socials: X https://x.com/leehomyc

Yue Zhao, Co-Founder & Chief Scientist

Zhao, the creator of PyOD—an acclaimed anomaly detection AI used in Tesla vehicles and NASA—has made significant contributions to the field of machine learning.

His open-source projects, which have garnered over 20,000 GitHub stars and 20 million+ downloads, reflect his impact on the industry.

Zhao's research includes influential work on ADBench, TDC, and TrustLLM.

He has received numerous accolades, including the AAAI New Faculty Highlights Award, Google Cloud Research Innovators recognition, Norton Fellowship, Meta AI4AI Research Award, and the CMU Presidential Fellowship.

A Professor of AI at USC, he holds a PhD in AI from Carnegie Mellon and has published over 40 papers on deep learning, ML robustness, and systems.

Zhao also serves as an associate editor for IEEE Transactions on Neural Networks and Learning Systems, an action editor for the Journal of Data-Centric Machine Learning Research, and is involved with KDD 2023 and various ML conferences.

His socials: X and LinkedIn https://x.com/yzhao062

Marco Di Maggio, Co-Founder & CSO

Di Maggio, the founding Director of the Harvard Fintech, Crypto, and Web3 Lab, is a leading figure in finance and technology.

Currently a Professor of Finance at Imperial College and formerly a Professor at Harvard Business School, he was named one of the Best 40 Under 40 Business School Professors by Poets & Quants in 2016.

He also serves as a faculty research fellow at the National Bureau of Economic Research and advises Coinbase, as well as various DeFi and TradFi companies.

His extensively cited research has appeared in top academic journals and has been highlighted by major media outlets, including the Wall Street Journal, The Economist, Bloomberg, CNBC, and Forbes.

Di Maggio holds a PhD in Economics from MIT.

His socials: LinkedIn

Patrick Colangelo, Co-Founder & CEO

Veteran in building technology at scale and a Harvard graduate, Colangelo seemingly does not have much presence digitally.

His X account's last post as a repost in February.

His socials: X

Xiaofan Liu, Co-Founder

Liu, a Lab for AI-powered Financial Technologies and a Professor at City University of Hong Kong, has a PhD in Network Science.

Like Colangelo, Liu also seemingly does not have much presence online.

Other than a short description on the team's page on their website, no other details can be found.

His socials: NIL

Nesa's Socials

Website

X

Discord

Is the World Ready for Decentralised AI?

Nesa's Layer 1 blockchain marks a groundbreaking advancement in blending AI and blockchain technologies.

By offering a platform tailored for AI, ensuring reliable proofs, and enabling critical on-chain inference, Nesa is set to become a key infrastructure for AI applications across industries.

Coinlive views Nesa's innovative approach as a transformative step in enhancing AI capabilities, ensuring they are delivered securely, transparently, and at scale.

Despite being in its early stages and lacking a LinkedIn presence, Nesa has already established partnerships with notable companies such as Sei, Privasea, XION, and MantaNetwork, bolstering its credibility and potential impact.

Yang hopes that future users will offer valuable optimisation suggestions and insights as they delve into Nesa.

He believes this feedback is crucial for Nesa's ongoing advancement and development, positioning the platform as a significant player in the Web3 industry.

In an interview in June with Foresight News, Yang noted (translated from Mandarin):

“There are three main things to look forward to in the short term: the first thing is of course the launch of the public mainnet, which will be accessible to everyone; the second thing is the integration of additional security and privacy features, such as zero-knowledge proofs, into more models, allowing users to enable or disable them based on their needs; and the third things is the development or hosting of more models on Nesa, which are expected to outperform and run faster than existing open-source models. Users will be able to use our models for tasks such as image generation, chatbots, and video creation. These are the key advancements we can anticipate from Nesa in the near future.

He added that the public mainnet should be launched in a month's time.