Real AI use cases in crypto: Crypto-based AI markets, and AI financial analysis

We are deploying genuine use cases for AI and cryptocurrencies every day this week, including reasons why you shouldn't necessarily believe the hype. Today get two for the price of one: blockchain-based AI markets and financial analytics.

It may not seem like the most interesting use case combining AI and cryptography, but both Near co-founder Illia Polosukhin and Framework Ventures founder Vance Spencer cite blockchain-based marketplaces that pull data and compute for AI as their best option.

AI is an incredibly fast-growing industry that requires increasing amounts of computing power. Reportedly, only Microsoft is invest $50 billion in data center infrastructure in 2024 just to meet demand. AI also needs huge amounts of raw and training data, labeled into categories by humans.

Polosukhin believes that blockchain-based decentralized marketplaces are the ideal solution to help collectively obtain the necessary hardware and data.

"You can use [blockchain] to build more effective and more equal markets,” he tells Magazine, explaining that AI projects currently need to negotiate with one or two large cloud providers such as Amazon Web Services. Still, it is difficult to access the required capacity due to a shortage of Nvidia's A100 graphics processing units.

Spencer also cites blockchain-based marketplaces for AI resources as his current number one use case.

"The first is to get real GPU chips," he says. “Where there is a huge shortage of GPU chips, how do you get them? [without] Really have a network that generates, provides and drives a market?

Spencer highlights Akash Network, which offers a decentralized compute resource marketplace on Cosmos, and Render Network, which offers distributed GPU rendering.

"There are some pretty successful companies that really do it at this point that are protocols."

Another example of a decentralized marketplace offering cloud computing for AI is Aleph.im. Token holders in the project can access compute and storage resources to run projects.

Libertai.io, a decentralized large language model (LLM) runs on Aleph.im. While you might think that decentralization would slow down an AI to the point where it couldn't function, Aleph.im founder Moshe Malawach explains that's not the case:

“The thing is this: for a user, all inference (when you generate data using a model) runs on a single computer. Decentralization comes from the fact that you access random computers on the network. But then, it's centralized for the time of your request. So it can be fast.”

Another blockchain-powered AI marketplace is SingularityNET, which offers various AI services (from image generation to colorizing old photographs) that users can connect to models or websites.

An emerging blockchain-based AI market that Spencer is very excited about is the tokenization and trading of AI models. Framework has invested in the AI ​​Arena fighting game, similar to Super Smash Brothers, where users train AI models that fight each other. The models are tokenized as non-fungible tokens and can be bought, sold or rented. "I think that's really cool," she says. "It's interesting to have native crypto monetization, but also ownership of these models."

“I think one day, probably some of the most valuable models, some of the most valuable assets on the chain, will be tokenized AI models. At least that's my theory.

Don't believe the hype: Currently, you can source components, data, and compute through traditional Web2 marketplaces.

Additional use case: Financial analysis

Anyone who has tried to interpret the ocean of data produced by on-chain financial transactions knows that while it is one thing to have an immutable and transparent record, it is quite another to be able to analyze and understand it.

AI analytics tools are perfectly suited for summarizing and interpreting patterns, trends and anomalies in data, and can potentially suggest strategies and ideas for market participants.

For example, Mastercard's CipherTrace Armada platform recently partnered with artificial intelligence company Feedzai use technology to analyze, detect and block fraudulent or money laundering-related crypto transactions on 6,000 exchanges.

Elsewhere, GNY.ioThe machine learning tool attempts to forecast the volatility of the top 12 cryptocurrencies and its Range Report uses ChatGPT-4 to analyze trends and buy/sell signals.

But can AI also help in traditional markets? That's the hope for Bridgewater, who will launch a fund next year from his new Artificial Investment Associate (AIA) Laboratory that aims to analyze patterns in financial markets in order to make predictions that investors can take advantage of.

Previous attempts to do this have produced mediocre results – with a Eurekahedge index of a dozen AI-powered funds underperforming its broader index of hedge funds by around 14 percentage points in the five years to 2022.

This is mainly due to problems related to feeding the large amounts of precise information that is required.

Ralf Kubli, a board member of the Casper Association, believes that AI can revolutionize traditional finance, but only if it combines blockchain records with rigorous standards to ensure that the information fed to the models is complete and accurate.

For years, he has been advocating for the financial industry to adopt the Algorithmic Contract Types Universal Standards, or ACTUS, created in the wake of the global financial crisis, which was caused in part by complicated derivatives where no one understood the liabilities or cash flows. involved. He believes that standardized on-chain data will be essential to ensure trust and transparency in model results.

“Basically, we believe that without blockchain, AI will miss quite a bit,” he tells Magazine. “Imagine that you are going to invest in an artificial intelligence company and every three months they update you on the progress of their LLMs, right? “If you can’t verify what they brought to the model, there is no way to know if they are making any progress.”

He explains that blockchain protects against companies manipulating their results, "and the past would indicate that […] “There is so much money that they will be wrong about what is happening.”

“AI, without this blockchain security layer (what happened, when, where, what was used), I think will not be effective in the future.”

He says the combination of the two will lead to new predictive capabilities.

“To me, the hope for AI in the future is that prediction models become much more powerful and behavior can be predicted much better,” he says, pointing to credit ratings as an example.

“AI used in the right way could potentially lead to much more powerful prediction models, which would mean that certain people who currently cannot get credit, but who would be creditworthy, can get credit. That is something that I am very passionate about.”

Don't believe the hype: AI's predictive capabilities have been shown to be poor at best so far, and reliable and trustworthy data that is not recorded on blockchain can be useful data for AI analysis.

Also read:

Real Cryptocurrency AI Use Cases #1: The Best Money for AI is Cryptocurrencies

Real AI Use Cases in Crypto #2: AIs Can Run DAOs

Real AI Use Cases in Crypto #3: Smart Contract Audits and Cybersecurity

Real AI and Crypto Use Cases #4: Fight AI Counterfeits with Blockchain