AI tokens: can they live up to the hype?

The recent success of ChatGPT has fueled the AI-themed token rally, with some coins posting triple-digit gains. Still, the community seems divided on whether it's actually possible to marry blockchain with AI.

Security officer watching cloud blocks forming face in sky - stock photo

Artificial Intelligence (AI) has been making headlines for years, but the trend had really blown up in late 2022 when OpenAI released ChatGPT, a powerful chatbot that can produce virtually any text-based content, including academic essays, programming code, fictional stories, and poetry. The platform was reported to have reached 100 million monthly active users in January, just two months after launch, which makes it the fastest-growing consumer app in history.

Quite naturally, many crypto enthusiasts immediately started looking for a possibility to combine two breakthrough technologies, AI and blockchain, in a quest for the next big thing. After all, this is nothing new to the crypto industry — we’ve already seen the rise of IoT tokens in 2019 when the Internet of Things was at its peak of the Gartner hype cycle.

Read also: Will quantum computing put an end to crypto?

So, what are AI tokens, exactly? As you could already guess, these cryptocurrencies leverage the power of AI to somehow improve decentralized governance, scalability, and security of the given blockchain. They can also be used as an incentive layer in AI-based projects, like generative art, AI-powered trading bots, or renting one’s idle computing power to train deep learning algorithms.

However, putting AI directly on the blockchain is currently an unachievable dream, Fantom founder Andre Cronje believes. The DeFi OG pointed out the fundamental differences in the underlying principles of artificial intelligence and blockchains, warning his followers against jumping on the AI hype train.

“Blockchain = Slow (by centralized standards), transparent, secure. AI = High throughput, opaque, blackbox. I need to very clearly stress, these two things do not mix. That’s like asking ‘What role do you see for Coca-Cola in the construction industry?’,” Cronje tweeted.

In this context, the “blackbox” refers to the fact that with most AI-based tools, the process of decision-making is not clear — researchers typically know the inputs and outputs of the learning algorithm, but it's hard to see what’s happening inside. This also means that if there’s a bias in the data, the model will spit out wrong predictions.

According to Cronje, “Blockchain and AI are not complementary, and (for now) anyone that jumps on the ‘AI’ bandwagon is simply doing so for pump and dump reasons, if you see a project all of a sudden ‘pivot to AI,’ it just means they had nothing and are dead in the water.”

“If (and that’s a very unlikely if) we get to a point where blockchains could handle the kind of throughput required that a neural net needs, we might be able to see them on-chain, but event then, the question would be, ‘but why?’. Blockchains don’t improve AI, and AI doesn’t improve blockchain,” the founder concluded.

“Let's make something clear. There is no material integration between AI and Crypto. You cannot put AI 'into a blockchain'. AI tokens are meme tokens,” Bankless co-founder David Hoffman tweeted, echoing Cronje’s point of view.

However, Ocean Protocol co-founder and AI researcher Trent McConaghy holds a different opinion on the possibilities of combining AI and blockchain. McConaghy believes that blockchain technology can greatly improve AI, especially in the field of decentralized data exchange, as data is crucial for training models but is increasingly hard to obtain.

“I get why you'd be annoyed by the AI, and AI * blockchain hype wagon. Looks similar to prior blockchain hype wagon. Just because there's hype doesn't mean that the thing being hyped isn't true! Blockchain benefits are real. AI too. And, AI * Blockchain too,” McConaghy stated.