The first wave of DePIN initiatives, around 2019, focused on digital infrastructure, but now we're seeing other types of networks emerging (DePIN stands for decentralized physical infrastructure networks). Projects centered on data or service networks are becoming more common. Ultimately, I classify DePINs as those that 1) use blockchain-based, decentralized coordination to operate infrastructure and 2) rely on or impact physical infrastructure like servers, sensors, or property.

Before moving on to the drivers, we need to understand that DePIN projects are almost always made up of two-sided marketplaces. These marketplaces have a demand and supply side.

Demand Side. The side where users look for services or products to address a specific problem or need, which can be met by a service or dApps.Supply Side. The side of the market where the decentralized infrastructure, including nodes, hardware, sensors and more, are hosted through the front-end of projects or dApps.So, what is happening on both sides of these marketplaces?

Supply side

Making it easier to supply DePIN allows for better, more diverse supply, which fits closer to the diverse demand side.

I see two main drivers behind the supply side coming online:

Cost curves coming downUtility token design improving

Cost Curves

Historically, hosting infrastructure required upfront capital that only large centralized entities had access to. With cost curves coming down, it’s feasible for almost anyone to become an infra provider.

Recent studies suggest that memory costs have dropped 100x over the past two decades, and compute (GPU) costs have decreased by 100-300x. Although demand for these resources is growing (and there’s even a shortage), the entry barriers for hosting significant compute or memory have drastically lowered. The capital required to build infrastructure is shrinking, enabling more people to participate, run nodes, and make the network more robust without key points of failure.

Utility Tokens

The design of utility tokens has long been seen as a dark art. Over the past years, know-how has been improving rapidly to deliver more robust token models.

DePIN networks often rely on utility tokens due to their network effect, as these tokens align incentives among stakeholders with different economic interests. Sound token design is critical for creating the right game theory and incentivizing behaviors that support the network, rewarding contributions appropriately. Utility tokens also help kickstart the initial network effect, and more token engineers are now using scenario analysis and statistics during design. This should lead to more robust designs capable of withstanding time and market volatility.

Demand Side

Historically, DePIN projects have been demand-constrained, meaning that services and applications were live but saw low take rates for a variety of reasons. Increasing demand finally makes DePIN businesses viable, kickstarting the flywheel of improvements.

I see three main drivers behind the demand side coming online:

Usability of DePIN is improvingPrivacy & security is becoming more of a concernData generation is exploding.

Usability

Let’s be honest: Many Web3 apps today are unusable to anyone who hasn’t spent significant time in crypto. Account abstraction and AI-enabled UX should solve this.

2024 is the year when account abstraction (which hides from users some of the technical wiring of blockchain transactions) picked up across Web3. The realization that the current Web3 user experience may not be compelling enough to persuade mainstream users to transition from Web2 has recently generated a lot of attention. Today, there are a large number of companies focusing on UX and account abstraction. At the same time, we’re seeing ERC-4337, the 2023 Ethereum upgrade focused on exactly this, really finding adoption with a wide range of projects, including DePIN.

Meanwhile, AI has been experiencing a renaissance since the launch of GPT nearly two years ago, with models improving and integration evolving rapidly. AI assistants, now being developed on blockchain, are set to simplify application use, reducing the need for human-friendly front-ends.

Privacy & Security

Concerns around data protection are proving beneficial for the adoption of DePIN, which foundationally improves this due to its decentralized nature.

While the privacy paradox is a well-documented reality, since the diffusion of AI across society, users are increasingly concerned. Particularly around data management, privacy and security are a growing concern. We see evidence that users are increasingly looking for alternative solutions that prioritize the protection of personal information. DePIN, with its decentralized approach, inherently enhances privacy and security, making it a more attractive option for individuals and businesses. The increasing sensitivity around this topic after decades of apathy creates a nice tailwind for DePIN.

Data generation

It’s estimated that around 350 million terabytes of data are created EVERY DAY. Humanity is generating an unprecedented amount of data that needs to be stored in memory and processed with computers, something DePINs are pretty good at…

We are creating more data than ever before. It is estimated that 90% of all data today has been generated in the past two years. With Gen AI, data has genuinely become the oil of the 21 century, so we need to make sure we store it well. Previously, many companies and individuals were agnostic whether it was stored in bare metal servers or the cloud; now there’s more of a decision process regarding data storage. With DePIN maturing, it’s finding itself more and more a viable alternative to the more established options for storage and handling of data.

What’s next

DePIN has a lot going for it, which explains the excitement among users, investors, and the entire community. I firmly believe that DePIN will soon redefine how economically important infrastructure is organized in society, positioning itself at least on par with traditional infrastructure.

Note: The views expressed in this column are those of the author and do not necessarily reflect those of CoinDesk, Inc. or its owners and affiliates.