AI infrastructure company Gradient Network based in Singapore has just completed a $10 million seed funding round, led by venture capital funds Pantera Capital and Multicoin Capital, along with HSG (formerly Sequoia Capital China). This investment will be used to develop pioneering protocols to distribute AI computing power across everyday devices.
The vision of decentralizing AI.
#GradientNetwork aims to build a decentralized AI infrastructure platform. According to Eric Yang, co-founder of Gradient Network, "We believe that intelligence should be a public good, not a corporate asset." This funding will help them build infrastructure that brings decentralization to the core of AI.
The timing of Gradient's funding coincides with AI companies facing criticism over data privacy and the centralization of computing power in the hands of a few tech giants. Gradient's approach is to harness unused processing power from smartphones, computers, and other devices, creating a global supercomputer powered by the community.
Two core protocols: Lattica and Parallax.
Gradient Network intends to use this funding to develop two core protocols – Lattica and Parallax – that will enable artificial intelligence models to run on a network of distributed devices instead of centralized data centers. Both of these protocols are expected to launch this week.
Lattica: Functions as a peer-to-peer data communication protocol similar to Bitcoin or Torrent. Lattica is a "pipeline system" that moves information between devices without going through a central server. The company states that their "Sentry Nodes" network has facilitated over 1.6 billion connections across more than 190 regions.
Parallax: Solves the problem of how to run massive AI models without needing large data centers. This protocol breaks down large language models into smaller parts that can run concurrently across multiple devices. Instead of sending data to OpenAI or Amazon servers for processing, Parallax will allow computation to occur on a network of participating devices, keeping user data local.
Gradient believes that this decentralized approach can significantly reduce costs compared to traditional cloud computing while addressing privacy concerns. When AI models run on centralized servers, queries and user data are transmitted and processed by those servers. Gradient's system will handle data closer to where it is generated. However, critics have raised concerns that coordinating tasks across thousands of devices could be complex, and network latency remains a challenge for decentralized systems.
Solana is the core platform and position in the decentralized AI industry.
Gradient Network operates on the Solana blockchain, chosen for its high transaction speeds and low costs compared to other networks. Solana will handle the coordination mechanisms and payments for devices contributing computing power to the network.
This startup enters a growing field of companies trying to decentralize AI infrastructure. Competitors include SingularityNET (focused on creating a marketplace for AI services), the Superintelligence Alliance network, and many other projects built on various blockchains. Bittensor and Gensyn are also pursuing similar distributed computing models, albeit with different technical approaches.
Gradient stated they will release additional protocols beyond Lattica and Parallax, although they have not specified what might be included. The company also mentioned plans to publish research papers and open channels for developers to contribute to the project.