The Magnetism of Ai and Blockchain

Ai and Blockchain are opposing technologies. One brings digital abundance; the other brings digital scarcity. Just as opposite ends of a magnet are violently pulled together; so too do Ai and blockchain seem inescapably connected.

The technologies

Ai refers to algorithms which make predictions. Ai looks at past data and uses that to generate new data. Popular examples include ChatGPT for generating text and Stable Diffusion for generating images.

Blockchain refers to algorithms which guarantee certain outcomes. Blockchains apply a pre-determined set of rules to future sets of data. Examples of blockchains in action are bitcoin and Ethereum.

In Bitcoin’s blockchain, the “future data” referred to new transactions. Some of the rule’s bitcoin’s blockchain implemented included not sending transactions twice, and having sufficient coins for a transaction. Whilst bitcoin’s blockchain was restricted to transaction data, Ethereum’s blockchain accepts more complex data sets contained in so-called “smart contracts”.

Both technologies have introduced automation. In Ai, we are seeing an automation of lower-level work; with the technology carrying out actions we would have previously given commands for. On the other hand, blockchain automates higher-level work; with the technology bypassing higher-ranking authorities. For example, with Bitcoin, we were able to send digital transactions without the supervision of banks for the first time; and with Ethereum entire financial ecosystems were conjured without any central intermediaries.

These two technologies are diametrically opposed. With Ai, we invented digital abundance: half of the content generated online comes from Ai. With blockchain we invented digital scarcity: bitcoin is widely referred to as digital gold for its mathematically finite properties. Being opposites, these technologies are also complementary.

Blockchain will unleash the potential of Ai

Ai chatbots already have higher IQs then human beings. However, we are only scratching the surface of what these new generative technologies can do. Whilst currently static, Ai will become dynamic. Ai agents will execute real world actions, taking control of our healthcare, our travel, and our finances.

In reality, Ai is already capable of doing much of this. However, Ai models are trained, owned and run by tech giants like Microsoft, Google and OpenAI. Therefore, society is hesitant to let these Ai models control our lives.

When Ai models are held under the control of tech giants, individuals are less sovereign because of the profit-driven purpose of these cooperations. Eventually, Ai will be able to predetermine intentions from brain waves and act on a principled basis. For such levels of power to be handled by Ai, the control must be kept with the individual. We need a mechanism that ensures these more advanced models allow us to retain our agency. This mechanism is blockchain; which will enable self-sovereign Ai.

Through blockchain technology, authority is decentralized. Just as we were able to decentralize money, we will also be able to decentralize the infrastructure of data.

Blockchain introduces a novel security infrastructure that can remove tech giants. Just as blockchains were used to send transactions without relying on a bank; blockchains will also be used to augment intelligence without relying on a tech giant.

By disintermediating the middleman, blockchain introduce self-sovereign security as each participant contributes to and owns a piece of the network. When user sovereignty is guaranteed; we will be less hesitant to unleash the full potential of Ai.

The coordination problem

Whilst the promises of predictive Ai technology transposed on top of blockchain security infrastructure is exciting. Many doubt that decentralized systems can produce the raw materials necessary to power the next generation of Ai. However, Bitcoin proved that crypto incentives work. The Bitcoin hash rate is 100 times more than the combined energy usage of Amazon, Google and Facebook. And as we enter into an era of ever more powerful Ai, compute as currency makes more and more sense.

Just as crypto incentives (underwritten by the compute of a decentralized network) where used to create sound money; crypto incentive mechanisms can also be used to create self-sovereign intelligence. That is, Ai models that serve humanity instead of serving the profit motives of tech giants.