Yield App

AI and blockchain: a marriage made in heaven?

7 min read

Last week, the intersection of artificial intelligence (AI) and blockchain made headlines once again after an Ethereum developer got ChatGPT to launch its own ERC-20 token. The developer asked the large language model chatbot to design and issue a token with an estimated market cap of $3.5 million by feeding it data from the 10,000 most traded assets on Uniswap. 

What emerged was the AstroPepeX (APX) token, which amassed some 2,300 holders and 17,700 transactions just one day after mint – although its price has been extremely volatile. This latest episode highlights the potential synergies between AI and Web3, and points to the possibility of merging these two innovations for greater benefits.

2023_Blockchain x AI 2.0-05.png

👰🏼‍♀️ AI: a copyright thief? 

Launching tokens for fun is not what it’s really about. Instead, blockchain could solve various issues that plague machine-learning technology. These include concerns over privacy and data protection, ethics, the authenticity and correctness of information, copyright issues, and more.

Machine learning isn’t quite the same as AI. Technically, AI is an umbrella term that denotes any type of machine intelligence that mimics humans, while machine learning is a subfield of AI that encompasses the process of training machines with algorithms and data to perform complex tasks. 

Yet, while these machines can undeniably perform calculations and answer complex questions much faster than humans, it’s difficult to ensure that their responses are correct and don’t violate any laws. For example, Meta’s LLaMA AI platform is facing a lawsuit in San Francisco from a group of writers who allege the platform “copied and ingested” their copyrighted work during training without their consent.

🤵🏼 Blockchain to the rescue

Attempting to verify the authenticity, correctness, and copyright status of such a vast amount of information is no small feat. This is where blockchain comes in. Using data audits blockchain can check for copyright infringement, accuracy, and even biases. Mrinal Manohar, CEO of Casper Labs, goes as far as to call blockchain “the world’s strongest copy protection technology”. 

Blockchain could also provide more actual data – especially in areas that are subject to strict data protections, like healthcare and finance. It can do so by incentivizing users to offer their data – though, of course, this must be done in an ethical manner to avoid the negative headlines that plagued Worldcoin following its highly-anticipated launch. 

A project that made a great deal of noise with its futuristic orbs that collect data by scanning people’s irises, Worldcoin drew criticism for the potential misuse of this data for marketing and other purposes. As such, strict controls and privacy rules are essential for an incentivization system to function, actively backed by regulators to facilitate such projects at a large scale. 

🧭 Pioneering projects

While some applications of AI may sound far-fetched, there are already several innovative projects looking to combine crypto with AI. These include decentralized data marketplaces like Ocean Protocol and SingularityNET.

The growing need for privacy could also be a boon for zero-knowledge proofs – a novel technology that allows one party to prove to another that a piece of information is correct without revealing the underlying data. This is the underlying technology in zk-rollups, a type of Ethereum scaling solution.  

LEARN MORE: An introduction to Layer 2 blockchains 

According to Marcello Mari, co-founder and CEO of SingularityDAO, this technology even opens up opportunities to “facilitate decentralized identity verification, while preserving privacy and anonymity”. He adds that large language models and generative AI can also “help understand tokenomic flows across different ecosystems”.

2023_Quote card blockchain AI_blog (1).png

🔌 Energy transition

Another important application for blockchain lies in the sustainability arena. After facing criticism for the high energy consumption associated with bitcoin mining, the crypto industry has responded by shifting toward renewable energy sources and utilizing excess energy to power mining farms. As such, blockchain can help AI solve its energy problem – and, in fact, some blockchain mining operations are already adapting their models to do just that.

AI’s energy problem isn’t a small one, either. Researchers estimated that creating GPT-3 required 1,287 megawatt hours of electricity and generated 552 tons of carbon dioxide – the equivalent of driving 123 gasoline-powered passenger vehicles for a year. Meanwhile, a report from the MIT Technology Review estimates that training an average AI model emits as much carbon dioxide as five American cars during their entire life cycle.

2023_Blockchain x AI 2.0-06.png


Generative AI models like ChatGPT are in fact, resource intensive, with a heavy reliance on GPU computer chips processing and cloud computing, which require both electricity and water for cooling. Today, cloud computing globally uses more energy than some entire countries. 

Projects like CUDOS, a decentralized blockchain-based network, are developing distributed cloud computing networks for Web3 and AI ecosystems. They aggregate resources from various data centers, powered by renewable energy, leverage idle computing resources, and reuse excess heat from data centers. This makes it possible to efficiently scale infrastructure while driving down consumer costs.

📈 AI and crypto trading

While blockchain has the potential to solve problems for AI, the opposite is also true. Notably, AI could greatly enhance trading. Already, between 60% and 75% of trading on major global stock markets is algorithmic, and a growing proportion of this is powered by AI. 

One of the advantages of using AI for trading is that machines eliminate the impact of human emotions. Emotional biases can negatively affect investment returns – in fact, there is an entire academic field dedicated to this called behavioral finance.

2023_Blockchain x AI 2.0-07.png

In the crypto realm, AI is already being used by multiple exchanges and blockchains for market, data and risk analysis. This includes Coinbase Global, which is using ChatGPT to map out the risks of newly added assets; GNY.io, which has a machine learning tool that can forecast the volatility of the top 12 cryptocurrencies; and Solana’s Omni crypto bot that can support crypto staking.

💍 Will they survive the seven-year itch? 

There are many exciting developments in the world of AI and Web3 that could result in a long-term union. From better access to data to solving sustainability problems, AI and blockchain can work together to enhance existing frameworks. 

However, it’s worth noting that large language models like ChatGPT are yet to prove themselves over time. Much like starry-eyed newlyweds, the path could unveil exciting possibilities, yet their track record wouldn’t be long enough for even the highest-risk hedge fund to consider an investment. As such, caution is warranted until all the kinks are ironed out. Until then, we’ll be watching developments in this area with bated breath.

Share:

Unlock the full potential of cryptocurrency and grow your digital wealth


Unlock the full potential of cryptocurrency and grow your digital wealth