Untitled

April 25 (Reuters) – What do you get when you cross cryptocurrencies with artificial intelligence?

A seemingly sentient bitcoin that codes itself in the style of Japanese haikus? Alas not, though you do get billions of dollars of trading in a new class of crypto tokens.

The machine mania sweeping the tech world amid the launches of bots like ChatGPT and Bard has reached the cryptoverse, with interest in tokens tied to AI blockchain projects surging.

Average daily volumes for the biggest coins including SingularityNET, Fetch.AI and Render topped $1 billion in early February, hitting a two-year high, according to data firm Kaiko.

AI-linked blockchain products cover a gamut of services including payments, trading models, machine-generated non-fungible tokens and blockchain-based marketplaces for AI applications where users pay developers in cryptocurrency.

“This is exciting, it’s one of the first times machine-learning applications are being brought on-chain in a big way,” said Eric Chen, CEO of decentralized finance platform Injective Labs, though he cautioned: “The digital asset space is no stranger to hype, speculation and overzealous expectations.”

So far, the investment returns are strong. The CoinDesk Indices Computing Index, which includes AI-linked tokens, has risen 60% this year with a significant spike in February as OpenAI’s ChatGPT saw a surge in usage.

While trading volumes retreated in March, they remain above the crypto sector’s long-term average, and many tokens have significantly outperformed bitcoin with year-to-date returns ranging from 150% to 780%, said Kaiko analyst Dessislava Aubert.

There’s also been increased investment in the sector, with examples including CryptoGPT, where users can sell their data to AI companies, which raised $10 million in funding this month.

Yet despite the strong returns this year, the AI-crypto sector remains niche – the combined market cap of CoinGecko’s AI-classified coins is $2.7 billion, dwarfed by the $1.2 trillion total crypto market.

Some projects may be riding the AI wave without a sustainable plan, with the relative newness of the space meaning winners will likely be few and far between, market players warned.

“There’s a place for AI and blockchain to see some synergy, but I don’t know how many of the current projects are using it well,” said Ryan Rasmussen, Bitwise research analyst.

“You have to look under the hood.”

Crypto tokens tied to AI projects have seen a rise in daily transactions with the biggest tokens adding significantly to their market caps.

CRYPTO AI: BIG HOPE OR HYPE?

The potential of AI-linked crypto apps has investors hoping they can sort through the hype to identify projects that can help solve some problems, drive more users to blockchain products and guarantee some solid returns.

“Some specific AI projects could actually end up being the ‘killer app’ for public blockchains,” said Pranav Kanade, portfolio manager at VanEck.

He separates the AI-crypto world into products likely to see near-term adoption as they solve immediate problems, and longer-term bets.

In the near term, the rise of decentralized computing networks could allow users with unused graphics processing units (GPU) capacity to provide capacity to other users that could be used for resource-intensive AI learning models, Kanade said.

Similarly, some industry watchers see blockchain-based marketplaces as offering an easy way for system developers to gain market share and smaller users to access new AI tech.

SingularityNET is one of the biggest such marketplaces and has seen the market cap of its token jump from $52 million to over $414 million this year.

Other potential long-term use cases include using blockchain as proof for distinguishing between AI and human-generated content.

Many investors are aware they may be in for the long haul, but are hoping a few runaway successes will compensate for the risk, said Todd Groth, head of index research at CoinDesk Indices.

“You’re investing in projects, many will not see the light of day,” he added. “You just need a few names that will do quite well.”

Reporting by Lisa Mattackal and Medha Singh in Bengaluru; Editing by Vidya Ranganathan and Pravin Char

Our Standards: The Thomson Reuters Trust Principles.

Opinions expressed are those of the author. They do not reflect the views of Reuters News, which, under the Trust Principles, is committed to integrity, independence, and freedom from bias.

Go to Source