In 2024, the FLock.io team launched new products, raised funding, cemented partnerships, published academic papers and resources, grew the community, and co-hosted numerous events.

It has been a year of phenomenal dedication and growth. It culminated in the official launch of our mainnet on New Year’s Eve, and geared up for the FL Alliance launch on January 10 2025.

Since we got started in 2022, we’ve been on a mission to revolutionise AI by making it private, secure, and truly community-driven. Here’s a look back at our journey this year and a glimpse into what lies ahead.

AI Arena beta launched in May

AI Arena is a Kaggle-like competition where users earn rewards by staking tokens to help train and fine-tune models for a given dataset. This beta brought Web3 a leap closer to traceable contribution and on-chain incentivisation for data owners, model developers and compute providers. 

Our community has created over 500 AI models since we launched on train.flock.io in May. The vision is to create specialised models for diverse communities. For example, the BTC-GPT reached 10k model calls.

Mainnet launched on Base and TGE for FLOCK token

On December 31, we announced the official launch of our mainnet on Base. This launch coincides with the Token Generation Event (TGE) for our native token, FLOCK, which serves as our utility token and debuted on Bybit.

FL Alliance applications open

FLock spent 2024 building FL Alliance ready to launch on January 10 2025, for distributed AI training of LLMs on consumer devices.

This decentralised system enables users worldwide to collaborate in the training and enhancement of AI models, using everyday devices like laptops, PCs, and tablets as training nodes.

Funding: investment and grants

On March 28, we announced a $6 million seed fundraise co-led by Lightspeed Faction and Tagus Capital. 

It didn’t stop there. On Dec 19 we announced the successful closure of a $3 million funding round, bringing the total raised to $11 million. The round was led by DCG, with participation from Faction, Animoca Brands, Fenbushi Capital, OKJ, GnosisVC, Bas1s Ventures, A41, and GSR.

In September, FLock was the sole AI infrastructure recipient of Ethereum Foundation (EF) Academic Grants Round 2024, to expand on the paper “Defending Against Poisoning Attacks in Federated Learning With Blockchain” in IEEE Transactions on Artificial Intelligence.

Partnerships with DeAI leaders

The DeAI ecosystem is stronger when it collaborates. That’s why in 2024 FLock partnered with Akash, io.net, Ritual, Foundry, Chainbase, and many more.

We collaborated with the Aptos Foundation to create a specialised LLM that leverages community-contributed Move code to enhance AI-assisted coding on the Aptos blockchain. FLock’s Aptos LLM outperforms ChatGPT-4o in generating Move-specific code.

In April, FLock launched decentralised training and hosting on io.net. Together, we form a complementary duo: FLock provides the collaborative training and fine-tuning platform, while io.net provides decentralised compute. We then partnered with io.net to launch a world-first proof-of-AI in August.

The integration between Akash and FLock.io enables users to easily train AI models on decentralised compute. Foundry, a leading node operator in DeAI, participated in the FLock incentivised beta to provide valuable feedback. FLock partnered with Ritual: an open, sovereign execution layer for AI, enabling seamless integration of AI into apps or protocols on any chain. Chainbase, the world’s largest omnichain data network, became a key partner. We kickstarted the journey with our FLock x Chainbase Text2SQL task on AI Arena.

Publications

Docs, whitepaper & litepaper

In 2024, we revamped the FLock docs on GitBook, including an in-depth tokenomics section.

We published a mighty whitepaper titled ‘FLock: Federated Machine Learning on Blockchain,’ as well as a litepaper. These encapsulate our goal: FLock, predicated on community involvement and a staunch commitment to data protection, is poised to spearhead the democratisation of AI ecosystem by using blockchain.

Academic papers

‘Defending Against Poisoning Attacks’ paper

FLock published a paper on the role of staking mechanisms in preventing malicious attacks in the journal IEEE Transactions on Artificial Intelligence: ‘Defending Against Poisoning Attacks in Federated Learning With Blockchain.’

Most existing FL approaches rely on a centralised server, leading to a single point of failure. This makes the system vulnerable to malicious attacks when dealing with dishonest clients. We address this problem by proposing a peer-to-peer voting mechanism and a reward-and-slash mechanism, powered by on-chain smart contracts, to detect and deter malicious behaviors.

‘Prospect Theoretic Integrity Preserving Alignment’ paper

At the Royal Society we presented the paper ‘Securing Large Language Models with Prospect Theoretic Integrity Preserving Alignment,’ participating in the Beyond the symbols vs signals debate. 

LLM training focuses heavily on maximising the likelihood of the ‘right’ answer based on their training data (the ‘log-likelihood’). While this improves accuracy, it also amplifies biases present in the data and makes the models vulnerable to misuse.

This paper proposes a solution grounded in prospect theory—a psychology concept that explains how people make decisions under uncertainty. Instead of solely focusing on optimising the model's accuracy, it takes a broader view, aiming to maximise the overall utility (usefulness and reliability) of the model. This makes LLMs more secure and robust against attempts to weaponise them.

Systematisation of Knowledge (SoK) paper

In December, we rolled out a Systematisation of Knowledge (SoK) framework. This initiative aims to illuminate the road ahead for DeAI. The framework is a comprehensive guide for developers, researchers, and innovators, featuring a detailed review of over 40 DeAI projects.

AI Arena paper

We published the paper ‘AI Arena: A Blockchain-Based Decentralized AI Training Platform.’ It explores our blockchain-based decentralised AI training platform designed to democratise AI development and alignment through on-chain incentive mechanisms. 

AI Arena fosters an open and collaborative environment where participants can contribute models and computing resources. Its on-chain consensus mechanism ensures fair rewards for participants based on their contributions.

Looking ahead: the 2025 roadmap

2024 has been a year of growth, learning, and community engagement at FLock. As we move into 2025, we’re excited to continue our journey at the intersection of AI and blockchain. 

FLock spoke at and co-hosted numerous events in 2024. Highlights included the AI Powered Summit, Ethereum X AI Confluence Day, RHECon conference, and CUBE Summit. We can’t wait to meet more of you at events this year.

Having just launched our mainnet and FL Alliance, we can’t wait to see where our team and community take us. Stay engaged, stay inspired, and stay with us as we step into an exciting future.