Community Cloud: Bob Miles is the Australian founder and CEO of Salad Technologies. He imagines a world where we can all share devices and create a ‘global mesh’ to solve the big problems of the world.
With all the recent metaverse hype, you might have missed another notion that’s become even more popular among industry leaders: ubiquitous computing. It’s not a new idea, but achieving it will be a prerequisite to approximating all that the metaverse could be.
The concept shouldn’t be confused with edge computing, where the commonly cited goal is to reduce latency by geographically situating servers close to a data source. Ubiquitous computing anticipates a world where billions of internet-connected devices may cooperate like a global mesh network. If that just sounds like the Internet of Things in fancier packaging, you’re not exactly wrong, but you’re missing the potential. We’re talking about a total paradigm shift.
As the performance and technology stack of consumer-grade computing hardware approaches parity with commercially available cloud servers, the average home computer will inevitably help power the internet, fuel technological innovation, and even offer its owner the chance to support favorite causes and communities – all from the comfort of the keyboard.
Concurrent advancements in consumer hardware and encryption have made it possible. The biggest hurdle? You’ve got to convince the people at both ends to trust each other. Crypto might provide the answer. One day soon, sharing your computer will be as commonplace as hitching a ride on Uber, or inviting strangers into your flat on Airbnb.
The Road to Community Cloud
When most people think of cloud computing – and that’s admittedly not often – they’ll usually picture a multimillion-dollar facility operated by the likes of Google or Microsoft. But before you know it, ubiquitous computing could bring the cloud right to your home.
High-end consumer GPUs and CPUs have attained a level of performance that rivals the expensive hardware housed in the world’s fastest data centers. A powerful gaming PC is already equipped with enough under-the-hood optimizations to tackle high-performance compute tasks like crypto validation, 3D rendering, engineering simulations, or even progressive climate modeling.
Consumer compute capacity is the theoretical processing power of the world’s home computers and devices. By some estimates, it has outpaced corporate-owned data centers by a factor of ten in just the past decade! And most of our private devices (besides the phones we use to play Wordle) sit inactive throughout the day, just waiting for something to calculate.
Even in 2022, only half of the world has Internet access. As more people come online, and new interoperable devices complexify our relationship with the web, today’s exponential data demand will skyrocket at an unprecedented rate. There’s going to be a real market for elastic and available processing power, bandwidth, and storage.
With home hardware gone pro, there’s no doubt we’ll see more networked applications attempt to harness idle compute resources for third-party use. The only question is how to capitalize on the moment in a way that doesn’t violate the end-user’s trust. I like to call it “computesharing.”
Computesharing Is New Again
We’ve already seen what’s possible when people voluntarily share the power of their devices. For twenty years, researchers from UC Berkeley invited private individuals to aid in the search for extraterrestrial life using their free [email protected] software. By the time the project wrapped in 2020, over 1.8M unique users had contributed to the hunt on a distributed compute network.
In the early 2000s, [email protected] famously borrowed processing power from thousands of idle PlayStation 3 consoles to conduct medical research. Over five years, those networked devices performed more than 100 million hours of protein folding simulations to unravel the genetic structure of some of the world’s most virulent diseases. More recently, the volunteer community donated countless hours to understanding the structure of the COVID-19 virus.
These projects are perhaps the best-known examples of distributed computing at scale. To participate in either one, users simply had to download a software client onto a private device. Both applications made use of latent processing cycles shared from inactive hardware. Yet despite their success, neither one truly solved the problems of trust inherent to distributed settings. They simply avoided them.
Because engineers had designed their software to manage unique workloads, researchers could confidently distribute compute tasks to anonymous users with only minimal safeguards in place. And since the end-user knew the software originated with trusted academic vendors, they could lend their devices’ compute power without fear of wanton misuse (even if their antivirus software threw a red flag).
Community Cloud: How Cryptography Unlocks Doors
Some folks are more than willing to loan out their private devices to support noble efforts. But for the vast majority of people, a worthy cause isn’t motivating enough to overcome their natural skepticism about unknown actors accessing their property. Nor is it practical for end-users to download an app to run every single compute job.
Thankfully, recent developments in cryptography have made it possible to safely distribute intensive compute tasks at a ubiquitous scale.
In 2020, IBM engineers published promising research on using a little-known encryption technology to process highly sensitive data (like medical histories) in a way that makes malicious intrusion all but impossible, even with physical access to the host hardware. The process is known as fully homomorphic encryption (FHE). It effectively obscures a piece of code while it’s processing so that no one can make sense of it until the job is done. The encrypted results are returned to the person who created the software to be unpacked.
Though it has seen little practical use since its invention in the 1970s, FHE is but one of several cryptographic protections that may allow researchers and tech innovators to run their software across broad networks of anonymous devices, without exposing user data or proprietary algorithms.
When you consider that consumer operating systems like Windows 11 also permit access to certain local machine resources – the bobs, bits, and cycles per second – in a secure service layer, it’s easy to forecast a future where the world’s idle computers could join forces to do practically anything.
Community Cloud: The Crypto Breakthrough
We can build a sharing economy on this new digital frontier. But it’s important to acknowledge that so-called “zero trust” compute applications actually require highly trustworthy systems that demonstrably minimize harm. I believe it’s equally imperative that participation be consensual, transparent, and valued as a real commodity.
To achieve truly ubiquitous computing, we need to codify the relationship between end-users and compute-job creators as one of mutual benefit. If we can devise trustworthy systems that fairly reward user participation, there’s no limit to the ways we could leverage our common compute capacity for good. And, we can build a more equitable internet in the process.
The growing adoption of cryptocurrency is actually a good predictor that all this will come to pass. Retail crypto exchanges are headlining the Super Bowl. This is despite the fact that most folks have never read a white paper. Fewer still could pick Satoshi out of a lineup! (That’s a joke.)
Through proof-of-work validation, blockchain ecosystems like Ethereum effectively demonstrate how built-in monetization models can incentivize supply-side network participation. The potential for block rewards can generally entice enough anonymous contributors to sustain a network.
It’s this attractive incentive that gets people interested enough to learn about the secure, yet “trustless” systems involved. Give someone a basic understanding of blockchain fundamentals and the right hardware, and you’ve given them all they need to cryptomine.
If you can reliably and securely distribute any kind of job across a network made up of any kind of hardware, all that’s missing is a powerful incentive for the people at the keyboards. Sharing economies based on interoperable cryptocurrencies could provide that missing piece.
Community Cloud: Looking to Ubiquity
It’s now possible to safely leverage consumer-owned computing resources, reward private individuals for their contributions to third-party workloads, and access the unique capabilities of any number of interconnected devices to accomplish colossal-scale compute tasks.
Anyone who has ever contributed processing resources to a mining pool heuristically understands how to cooperate with unknown actors toward a common goal. Within a few years, I think we’ll see computesharing become as commonplace as ridesharing, flatsharing, or any of the other radical notions that now come pre-installed on your mobile phone.
About the author:
Bob is the founder and CEO of Salad Technologies. After starting his career as a pilot and aeronautical performance engineer at Qantas Airways, Bob co-founded the digital production company responsible for The Green Way Up, a twelve-part television series commissioned by National Geographic and distributed through Netflix. The program followed Bob and his fellow engineers as they journeyed across the Australian continent in a vehicle powered by a waste-to-fuel system of their own design. Bob went on to leverage several years’ experience product-managing mobile applications as the co-founder of a connected car startup that developed networked software for consumer automobiles. He later relocated to the United States to take a position as Head of Product at a consumer drone manufacturer, where his passion for aviation coalesced with his expertise in building networked applications.