blockworks.co
It’s funny, and even a little bit annoying, how a stale tech cliche can make a comeback. Just a few years ago, the mantra “data is the new oil” could give you a headache. It carried the weight of the endless strategic meetings, overly-enthusiastic sales calls and pitches, millions pumped into “data transformation” — a concept as alluring as it was vague.
And yet, here we are: The Big Data market is blooming, expected to blow up to more than $624 billion by 2028. From walled-off APIs to the Musk-Microsoft spat and Stack Overflow’s release of Overflow AI, a lot of the things that make the buzz in the tech world come down to one theme: data, and all the intricacies around it. Come ChatGPT and its AI kin, and the elden prophecy suddenly begins to make sense.
Turns out data is the new oil after all — but who’s holding the drill? Thanks to the most exciting development in years, going by the least exciting acronym in years – DePIN, the answer is you.
The DePIN model enables us to take ownership of our data and earn from it, or opt out of selling it if we so choose. It makes people stakeholders and players in the data market, not just data points to be harvested. In a sense, it takes drilling away from centralized giants and hands everyone their own personal oil rig, which is a lot fairer and more compliant.
Our robot overlords demand your seventh-grade essay
Maybe it wasn’t your best writing, but yes, SkynetGPT still wants a read. How else would it know how to impersonate a seventh-grader? And while we’re at it, it also wants to scrape every marketing blog out there, take a very close look at the latest financial forecasts, and check up on any copyright-free novels. Not that it understands anything (yet), mind you. For SkynetGPT, all these texts are nothing but data, which it processes to train itself to approximate human writing.
Data is the fuel powering the engine of machine learning, which produces both advanced generative AIs like ChatGPT and StableDiffusion and simpler models. And, as any fuel, it first has to be extracted and processed. Where from? Well… More often than not, from the lived human experience. Social media and forum posts, publicly-accessible texts and image archives, platforms for artists and other creatives — that’s where you drill if you need to train an AI. And those rigs better be pumping non-stop, cause you’ll need troves and troves. To put things into context, Stable Diffusion was trained on 2.3 billion images, which is a hefty amount.
Almost a decade ago, Google was forced to shelf an AR glasses project due to privacy outcry — where is the same outcry over Apple Vision Pro? We’ve grown used to Big Tech privacy intrusions, and the age of AI will put them on steroids. Your Reddit posts, your Instagram pictures, your Artstation displays — everything will work as the fuel for the AI engine.
And it’s not stopping there.
Data makes the world go round
The truth is, the rise of generative AI didn’t magically turn data into new oil — it has been that all along. ChatGPT and other generative models are stealing the show, but data also fuels a ton of other services we are using every day, from GPS navigation, which use live traffic data to find routes with the least amount of traffic, to weather services, which obviously want to know the latest temperatures, humidity, etc, in any given city or region.
By the same account, more and more devices we use every day work as data harvesters. Take the seemingly innocent fitness trackers — their makers usually deny they sell your data without your permission, but have you really read the terms of conditions on the app when signing up? Because that data is lucrative, and they could in fact be selling yours. The same goes for smartphones, smart home devices, office tech and more. The more devices we use, the more data gets captured and, often, sold off to third parties, without us getting a dime or even knowing about it in the first place.
Read more from our opinion section: Web3 is for regular people, too
But as we are nearing the always-connected dystopia brimming with sensors recording everything, there is a better, more democratic way of doing things. It’s called DePIN, short for Decentralized Physical Infrastructure Networks, and they are the engines powering the data markets of tomorrow in a more equitable way.
Everybody gets a drill
DePINs use tokens to incentivize people and businesses to direct physical infrastructure — i.e. devices, vehicles, or machines of any kind — to provide real-world services and earn. 5G networks, weather data, car sharing, street mapping, pollution monitoring and so much more — all on the blockchain, all peer-to-peer, all crowdfunded, crowdsourced, crowdowned.
Under the logic of today’s user data monetization, which Shoshana Zuboff aptly called Surveillance Capitalism, a human life is an oil deposit for Big Tech to harvest. DePINs turn this model upside down, enabling anybody to set up their own rig and opt to sell their data, properly anonymized.
From heath tracker data to insights gathered from AI-powered cameras or weather data from across the globe, the DePIN model can deliver any sort of data. It has already proven its ability to scale at a pace far beyond the capabilities of tech giants and deploy fast anywhere and everywhere in the world. So while the AI models that will power anything from mobility to weather prediction will require a lot of data, DePINs will surely deliver. DePINs aren’t just solving the problem at the end-user level, they’re replacing the whole system.
So what, let’s get drillin’?