In 2011, Marc Andreessen said something that was already eerily obvious to most people in the know, but had yet to be elucidated so clearly. He said, "
What Andreessen meant by his statement was that most businesses would soon come to depend on software to perform their most basic functions. He meant that taxi-hailing companies would soon come to rely on apps. Supermarkets would come to rely on apps. Even newspapers would come to rely on apps. He meant that every business would need software to be viable.
Roughly ten years later, Andreessen has been proven right. Every product now needs an app or software to be viable. However, as incredible as this idea of software-eating businesses is, it's stale news. There's a new beast in town, and that beast is eating every single thing, including but not limited to software.
That beast is called artificial intelligence, or AI for short. Over the last ten years, development in computing and machine learning has guaranteed that AI will be an incredibly important part of the software of the future. And the infrastructure underpinning Web3 will be no different.
In many ways, inevitably, AI will eventually "eat" software, just as Andreessen said about software eating businesses all those many years ago.
The question that remains now is not whether AI will be a huge part of Web3, but how Web3 will incorporate AI. It's a question of when not If. And we need to start talking about the How right now.
The How Of AI in Web3
There are three main components of a Web3 future. The first, and the most important part, is the blockchain itself. The second is the Dapps, which are the applications that run on the blockchain. And the third is the protocols.
In Web3, AI has the potential to be applied uniquely to all three aspects of Web3. And here is how.
The Blockchain
We'll start with the most important part of Web3, which is the blockchain. The key feature of the blockchain is the decentralized nature of Its computing. This decentralized nature means that it is near impossible to commit fraud on the chain since the records of every transaction are recorded on different servers.
But this ability isn't
In fact, something of the sort is already happening. In
Asides from that, AI will also be able to do other things for the blockchain. For example, an AI protocol may be able to predict transactions, thereby creating consensus protocols that can scale quite easily.
Dapps
AI will most likely figuratively eat Dapps first, and it is already doing. We are already seeing this trend with many Dapps on the market. For example, there are already Dapp games on the blockchain that make use of AI. In the future, we might even see games that will be able to change scenery, difficulty level, and missions following the mood of their players.
And this isn't some future possibility — it's already happening. __In 2019, __Cortex Labs announced that it had launched the first AI dApp unto its mainnet. The dApp was called Digital Clash, and at the time it was the first dApp to incorporate on-chain execution of machine learning models. Interestingly, the Cortex blockchain allows AI devs to upload trained models to the storage layer, and dApp developers can use these models for their own dApp development.
NFTs, to an extent, are also good examples of how Web3 will incorporate Dapps and AI. In the future, we might even have NFTs that will change how they look to reflect the mood of their users.
Again, all of this is hypothetical, but it's an interesting thing to think about when talking about the different ways Web3 can incorporate Dapps.
Protocols
Protocols are perhaps the one part of Web3 that is already being eaten by AI. Smart contracts and Defi seem to be ripe for AI-led disruption, and one can imagine that that would happen anytime soon.
Another part of DeFi that would probably see AI disruption first is loans. Imagine a world where an AI could determine your DeFi credit score by going through your online data. Imagine a world where that credit score is extremely accurate, and predicts — as much as a credit score can, how likely you are to pay back a debt.
Such an application, for one, would take the training wheels off DeFi. It would also spell an end for over-collateralization, which is a
The When
Now that we know the hows of AI and Web3, it's important to understand the when. Progress on AI systems is accelerating as fast as possible, and with every day, people are building exciting products in Web3.
However, this doesn't mean we know when Web3 will become viable enough to be mainstream. For now, most people are comfortable with centralized data, control, and apps.
But that will not last forever. As many are already realizing, it's dangerous to leave so much power with central authorities as they can pull the switch on any network at any time.
The important thing is that there are already platforms providing people with the tools to build for that future today. The best example of one of such platforms is
Web3 is already coming. We know the how of it, we understand the why of it, and we will soon know the when of it.