Internet Computer Teases AI Contract Engine Update

Internet Computer Teases AI Contract Engine Update

Internet Computer Teases AI Contract Engine Update

Internet Computer has made significant advancements in integrating blockchain-based artificial intelligence (AI) through smart contracts.

Almost exactly one month after discussing the launch of the world’s first blockchain-based artificial intelligence (AI) expressed on a smart contract, the Internet Computer (ICP) has now hinted at a significant advancement in the pursuit of this goal.

Internet Computer Is Advancing AI Smart Contract Capabilities

ICP founder Dominic Williams highlighted the tremendous acceleration of artificial intelligence smart contracts on the protocol in a piece that was published on X.

He acknowledged the work that the DFinity team had done on the smart contract engine, emphasizing the fact that it had brought about a tenfold acceleration in the execution of deterministic floating point instructions.

Williams stated, “We are continuing on our mission to create sophisticated LLMs that can function as smart contracts.” Image categorization is now operating three times faster in the demonstration (across three to four blocks through consensus, as opposed to ten to twelve blocks before).

It is only natural that the gas and cycle costs of artificial intelligence inference have decreased proportionally. The team was able to achieve this enhanced speed through extensive low-level work.

This work includes “commitments to the public Wasmtime virtual machine implementation of WebAssembly.” During the upcoming optimization, it is anticipated that the protocol will advance even further than it is.

It is important to note that ICP is devoting a significant amount of resources to its artificial intelligence (AI) endeavor. This brings the company closer to competing with industry leaders in AI, including OpenAI, Google, and Cognition Labs, home to the first AI software engineer, Devin.

More Features To Come on ICP AI Innovation

The Internet Computer had previously discussed the possibility of incorporating SIMD instructions into the environment in which smart contracts are executed, and this subsequent optimization will include the implementation of such a procedure.

As a result, in the long run, this will make it possible to carry out numerous floating-point calculations simultaneously by executing a single instruction. Additionally, there are arrangements in place to transition the ICP smart contract environment from 32-bit Wasm to 64-bit Wasm.

An important aspect of the migration is that it will increase the number of AI model weights that can be placed into contract memory. This number will increase from 2 billion to whatever is required for advanced LLMs, according to the founder of ICP.

The ICP team will also develop new application programming interfaces (APIs) that will enable canister smart contract code to export AI computations for accelerated processing on graphics processing units (GPUs).

These objectives show that ICP is committed to enhancing its artificial intelligence infrastructure. It was essential to introduce the Internet Computer AI in March to accurately convey the protocol’s objective, which is to establish a system in which powerful LLMs run as smart contracts on the Internet Computer Protocol (ICP).

The Internet Computer AI functions within the ICP testnet of DFINITY, enabling developers to transfer their code to the public network for execution.

Read Previous

BlackRock Uses Bitcoin as US Inflation Hedge

Read Next

Hedgey Protocol Hit by Cyber Attacks, $44.7M Lost