In the next twenty years, computers — mostly in the form of personal devices — will change society, culture and our lives. We believe the scope of this change will be like the Copernican revolution; it will signal a shift away from centralized economies, and towards an artisanal world where individual creativity, personal connections, and human-scale sharing become the central values.
To enable and empower this profound shift, we here at Mylio are planning a broad framework and platform called Gaia. The Mylio photo app we use today is the first step in that direction. Here’s a sneak peak at what we’ve been thinking about.
Computers have been around for over 60 years, marked by occasional technology ‘revolutions.’ Terminals, minicomputers, PCs, the Internet, smart phones — with each quantum leap, revolutionaries have hoped a less centralized, more democratized world would emerge. People used to worry about the dominance of IBM, and the ways in which mainframes amplified a world of centralized control. Then, we worried about Microsoft and its apparent control over the PC world. Today, ironically, as computers have grown smaller, more ubiquitous, and unimaginably powerful, centralized control has only become more of a concern. Google knows more about billions of individuals than most governments. Apple, Amazon and Facebook are all strong instruments of a future of centralized control, and pervasive reach.
But it doesn’t have to continue this way. Because the biggest technological change of all is underway. We’ve seen the shift from thousands of mainframes, to millions of PCs, to billions of phones — and we’re about to see trillions of even more powerful devices become central to our lives. As that happens, we have a once-in-a-lifetime opportunity to be sure the next technology shift leads away from centralization and de-humanization, and towards a world we all really want.
That’s what this paper is all about: societal change, cultural change, technical change; and most of all an imperative change. Imperative that is . . . if we care about the value of individuality and humanity.
The Great Acceleration: Processing Power, Storage, and Moore’s Law
The new NVIDIA chips in cars, and the Qualcomm chips in drones, completely change how we view ‘devices’ and ‘gadgets’. A mobile phone today is literally a supercomputer. What’s more amazing: a car is now a mobile phone on wheels, while a drone is a mobile phone with wings. New cars have local networks connecting over 100 CPUs; a self-driving car will have many hundreds of CPUs. The Nvidia AI car showcased at CES 2017 has 2 discrete processors, and 2 discrete GPUs delivering 24 trillion operations per second, drawing just 30 watts with 8 high-end CPU cores with 512 next generation GPI chips inside. Ten years ago, this kind of power and capacity weren’t even imaginable.
All these cameras, cars, phones, drones (and even Apple’s new AirPods) have domain-specific, vertical applications running on them, collecting, processing, and storing real time data. These supercomputers are everywhere, and even the smallest, most ubiquitous among them are quite amazing and powerful.
Consider that over one trillion phones have been produced in just a few decades, and we’re just getting started. Now that’s scale. But it’s the power of these devices that’s truly astonishing. A few years ago it looked like Moore’s Law — which predicts computers becoming twice as powerful every two years — might end. But we’ve been experiencing 120 years of Moore’s Law, and as the graph below shows, GPUs (Nvidia is responsible for the last 7 points) are keeping the Law as relevant (and revelatory) as ever. Add molecular and quantum technology to this mix, and we should prepare ourselves for mind-blowing increases in power and memory for the foreseeable future.
How small can an amazing device be? People love AirPods, but what’s behind the love? These tiny devices use machine-learning algorithms thanks to a new W1 chip that also, for the first time, makes Bluetooth connectivity totally transparent. So AirPods are both personal assistant and DJ, small enough to fit in your ear. And to make that happen, each Airpod has more computing power than the iPhone 1, with a total of 3 Cortex SoCs.
Today — or tomorrow at the latest — each of us will have our own constant, private, local cloud. Snap Spectacles may be sunglasses, but they can also record our lives, always and everywhere. Our phones, tablets, laptops; those intelligent AirPods, our cars, our cameras, that self-driving car, security cameras, baby monitors and even the intelligent door bell — all these exist for our personal use, and all part of a cloud that can be local, private, and ours alone.
The average person at the turn of the century might take 5,000 pictures in their lifetime. Today, it’s easy to take that many pictures in a week or a month. Over a trillion pictures were taken in 2015. Cars now generate 25 Gb of data per hour. A self-driving car generates around 10 Gb data per mile. A Lytro camera generates 300 Gb of data a second.
Now consider that there will be billions of humans, each of whom will be assisted and accompanied by dozens (or hundreds) of these smart devices, all which will be generating data, and all which will be expecting to interact with ‘their’ user. It’s clear that we’re on the verge of a simple, but profound sea change: intelligent, powerful devices, receiving and generating massive amounts of personal data, are becoming part of the fabric of our world.
We believe that making this world manageable and real requires a sea change in our computer architecture. How do we — how do you — collect and organize all that data?
The answer lies in how we answer these questions:
- Where’s the cheapest, most efficient storage?
- Where’s the cheapest, most efficient computing power?
- What’s the most agile, efficient way to deliver personalized machine ‘learning’ about a person’s own data?
And answering these questions tests some fundamental assumptions we’ve been making about computers.
Cloudy Today, But Not Tomorrow
Where’s all the ‘interesting’ computing done today? Where’s most of our personal data stored? Where do we go whenever we have a question, send a message, catch up with friends, research a product, or buy something? What’s been the single answer to almost every architectural question in the last ten years? Where should I store my photos? Where should I keep my documents?
This is a supreme irony. Because that
phone supercomputer in your pocket is 5,000 times more powerful — with 5,000 times more storage — than 60s-era mainframes. And yet, when your supercomputer connects to Amazon, Facebook, Google, iCloud, or Snap, it becomes a terminal. A handheld terminal, granted, but still a terminal.
In the 70s, terminals were supposed to democratize computers (they didn’t). In the 80s, the torch was passed to minicomputers, then PCs, and finally, the Internet. (They didn’t either). In each case, the original vision of computing by and for individuals turned into yet another case of encroaching centralization.
For the first time though, we have a technology change so big, so pervasive, and with such deep implications, that a centralized solution may not even be completely possible. If billions of humans each have dozens or hundreds of devices that they interact with every second and every minute, how does a centralized solution even begin to keep up with — let alone stay ahead of — that power curve?
In the paragraphs ahead we lay out the distributed, personally-focused world that we believe the Internet of Things calls for. The technical arguments are compelling, but you should also think about the personal arguments. The opportunity for an empowering change is in front of us. So what do we do about it?
The Future: Distributed At Last
1. Devices with sensors, computing and storage power — the drones, the AirPods, the phones, the cameras and cars, etc..
2. DIY personal mesh clouds: these personal devices — all of them — will form their own hyper-local mesh cloud for you, and you alone.
3. Hybrid Networks: Your personal cloud will connect, seamlessly, to the Shared Cloud. From you perspective, this Shared Cloud will be just another device. The Cloud will also be the place where your devices check in for updates, and generalized machine learning.
4. Privacy (At Last):
Our personal cloud finally puts all our information at our fingertips. What we share is then up to us.
5. Mobile Native Creativity Tools: Office, Creative Cloud and even the gSuite were created for knowledge workers and personal computers. But of the 5.4 billion people that have phones today, only 1.3 billion have computers. So tomorrow’s platform will include simple, yet powerful creativity tools that work equally well on desktop and mobile devices.
6. The Artisanal World: We see this phenomenon most clearly in China: the dividing line between individuals and small business simply does not exist. Even today, over 90% of the jobs in the world are in companies with fewer than ten people. As permanent, structural unemployment becomes a more and more pressing reality, crafts-based economies become a categorical imperative. In this Artisanal World, being able to be a first class citizen in the global information network is a matter of survival.
Suddenly, each world citizen, each with their own personal army of computing devices, is a force to reckon with. So long as we build the distributed architecture to make it possible.
7. Programming for People: Writing code is too hard. Like the Cobbler’s children with no shoes, billions of supercomputer users can’t even imagine specifying the behavior of the devices in their hands. Solving this problem is about languages, tools and engines. With Mylio, we start the process of putting a database on your phone — for the first time, a lifetime of data and memories can be with you always. Unlike corporate databases though, this one speaks to you in your own terms: family, business, events, places and more. That’s the engine, but the tools have to be there too. For the first time, programming will become possible for ordinary people.
8. Community: What does the this even mean in this new world? Society will be structured differently. It will be the re-birth of the artisan class, who will be commercially connected and integrated. They will empower each other with commerce and learning. Mentorship will be important in education and training. New forms of commerce will support barter and local exchange, bitcoin and value-based commerce. The community of local artisans, connected all over the world by the new platform, will be as powerful –and as global — as the largest corporation, and yet individual at the same time.
From Mylio to Gaia: Making It Happen
Amazon started with a bookstore, then grew into dozens of global platforms. We start with Mylio, and plan to grow it into a set of platforms we call Gaia. To see how this happens, let’s start with a simple architectural progression.
At heart, Mylio is a data platform. Today, Mylio users would describe it as a photo management system (and there are certainly lots of photos out there).
As such, it both starts the transition to Gaia, and provides us with critical learning about how to make this all happen. Before looking at Gaia, let’s drill just a little more into the Mylio architecture.
The Cloud Was Just the Beginning.
The diagram at left shows the centralized model that every photo service — Google, Apple, Yahoo and the rest — uses. You put your photos in the Cloud for syncing, for replicating, for storage, and for processing power. In this model, the Cloud is the hub, and your devices are the spokes. There’s nothing wrong with this model; it works . . . for now.
Because even with just photos, there are challenges with this cloud-only model. Constant connectivity is great — until it’s not there. And how comfortable do you feel with Google tending all your photos, literally holding the story of your life? Most all though: what about those supercomputers in our pockets?
Why Mylio, Why Now (and Beyond).
This is the bet we’re making at Mylio: building the first personal hybrid mesh network. We’re already facilitating data transfer, replication, and syncing via agile, efficient, private, peer-to-peer networks using the devices you already have, including storage devices like USB and NAS. Any file can be found in seconds. All your pictures are with you always. If you choose to also put them in the cloud via Mylio, that’s your choice — but if you choose not to use the cloud, you have that choice too. Either way, everything is everywhere — on all your devices — always. With or without the cloud.
In this model, We, the Humans are returned to the center of data flow. The Cloud becomes just another storage option, or the place where machine learning takes place, with new data being pushed back down to the devices. Data management and ontological work happens in the background; organization happens through a blend of machine learning and user input.
And in this model, a measure of economic power is returned to the people. Using devices already at hand empowers individuals and small businesses, who enjoy greater value from hardware in which they’ve already invested. But now comes the really interesting part:
Photos are just the beginning.
As of last year, Mylio handles documents too. Now I can have 100,000 pictures and 15,000 documents with me at all times on my phone, my tablet, my computer. Always. Anywhere. Even if I’m on an airplane with no connection, every picture I ever took, every presentation I ever made, every paper I’m reading, need to read, or have read — they’re all right there.
This coming year, we plan to build mobile native productivity tools. I can use my phone, my tablet, my computer, everywhere, anywhere, to create an invitation, edit a catalog, build a presentation, design a book. And, unlike the digital tools of the last century, these tools are powerful but simple, as only the mobile native world could have taught us. Add workflow and creativity tools to the mix — for editing, filtering, sharing, creating documents, presentations, graphics, or collages — and new worlds of art and commerce become attainable, greater impacts made possible with fewer resources.
But the implications of this paradigm shift aren’t just technical or economic — they’re philosophical too.
When we talk about Cloud services and companies like Google, we can start with the elephant in the room: information asymmetry and privacy. The hard fact is that Google has more information about you than you do, and while many people aren’t happy about this, most feel powerless to change it. Starting by taking your information off of someone else’s servers, and keeping it within your own private device network is the first step in such a change, an act of agency and empowerment. In this new paradigm, you can have a choice again for what happens with your data, and how. Mylio can help you get started.
Mylio places us firmly in the distributed, decentralized device world; how does it extend into Gaia?
Gaia is a set of products, tools, platforms and engines; Mylio is the first. Gaia has three missions:
1. Programming For People: A runtime, distributed platform and set of design and development tools that allow small businesses and ordinary people all over the world to build and run their own device and cloud centric-distributed applications. Really. It’s about time.
2. Community: Redefining the concept of community so that artisans, small businesses, and individuals all over the world have a decentralized platform for creating their own communities, sharing knowledge, and cooperating on projects. Community systems exist today, but they’re all centralized, and focused on large organizations. We want to change that.
3. Collaborative Commerce System: Digital currencies and barter frameworks along with micro-lending and community risk-sharing have the potential to provide a framework that supports and empowers the artisanal community of the future. A key goal of Gaia is to invent and build that framework.
So What’s it All Mean?
Given the number of devices and users, and the volume, complexity, and importance of the data we’re discussing, this next shift to distributed computing will be tectonic, with huge implications. Networking, storage, security, management, and programming languages will change. A whole new set of services and applications will help you create your own local Cloud, servicing your own devices. Programming will be match-based. It will start with vertical, specific problem solving. It will be artisanal.
These things will in turn change business models in ways we can only begin to predict.
We will be making more bets along the line, and building companies to maintain momentum. From architectural changes, to the opportunity to be the ontological shell, to innovating new database and data structures, to providing tools to create economically-connected communities and coin-based business models, to supporting distributed architectures and barter mechanisms… Mylio is here, and here is just the beginning.