In halcyon days of yore (circa 1995), computers always did what you told them. They were a delight to use. Ever since I was a kid and my parents brought home our first Mac, I've loved personal computing and the idea that I could create anything I could imagine.
There were no subscriptions, no login screens, no cookie banners, no notifications, no pop-ups, no up-sells, no ads, no rage-bait, and none of the crap to which we have grudgingly grown accustomed.
But Personal Computing 1.0 was far from perfect. It was clunky. It was under-powered. Your computer couldn't really talk to my computer (except by literally calling on the phone and making noises at each other). Information had to be accessed by finding the right floppy disk for the correct Encyclopedia Britannica volume.
Then The Internet happened.
Suddenly, all of humanity's collective knowledge was available at our fingertips. There were thousands--then millions--of incredible personal websites and blogs and new fun crazy experiments. A billion flowers bloomed.
But over time, those blooms became surrounded by noxious weeds. The ever-grasping slime-mold-that-is-Moloch reached its enshittifying tendrils around the roots of the web, slowly strangling it. Bit by bit, abandoned personal blog after dead side project, Big Tech shed it's "Do No Evil" skin and relentlessly "improved" the "efficiency" of our shared digital commons. Spammy listicles and product pages replaced "unmonetized" information in search results, and social networks shifted from "creating connection" to manufacturing addicts.
In a sense, this transition culminated with the rise of closed AI. These companies stole practically ~every book ever written and got off with a slap on the wrist (unlike Aaron Swartz, who stole far less and then actually died). They then went on to scrape all of the collective wisdom of the web (from Wikipedia, Stack Overflow, Reddit, Arxiv, etc), and now sell it back to us and while simultaneously arguing for laws to prevent anyone else from participating in the human exchange of information (aka, culture.) The primary net effect of all of this? A "dead internet" filled with endless AI slop--slop writing, slop images, slop music, and security-vulnerability-ridden slop code.
But history repeats itself.
Much like the mainframe computing era gave birth to desktop computing 1.0 as processors and memory became faster and cheaper, so too could today's closed AI labs give birth to an era of ubiquitous, smaller, open models. Everyone could have access to effectively unlimited intelligence and productive capability, and for marginally more than the cost of the electricity.
This could be the dawn of Personal Computing 2.0.
In my vision for this future, Personal Computing 2.0 is a computing paradigm that appropriately places AI (and all technology) as subservient to human concerns since, after all, the purpose of technology is to make our lives better.
Like Personal Computing 1.0, Personal Computing 2.0 puts the individual user at the center of the technological system. Personal Computing 2.0 is about making the interfaces and software environments that empower people to do their life's work. It's about unleashing the creative potential of individuals.
To accomplish this, Personal Computing 2.0's central principle is that your data is yours. From this, everything else follows.
The revolution begins with liberating your data. Instead of having your data "live in the cloud", i.e. where some company can force you into a subscription plan to access your old files or remove API access to your own data, Personal Computing 2.0 is about giving you full ownership over your data. You can use it however you want. You can take it anywhere you want.
Personal Computing 2.0 is a recognition that computing has become a central part of our lives. Losing my access to my Mac as a kid would have been sad, but ultimately forgotten. That is no longer the case: getting locked out of your Apple or Google account can mean losing access to most of your personal information, your photos, and even your personal memories. And being cut off from computing completely is almost unthinkable--many of us wouldn't be able to work or live without access to any computing devices. The ability to delete your life and career is far more power than any one company should have.
Thus, Personal Computing 2.0 must be cross-platform. It's inherently multi-vendor. Just like you could use Microsoft Paint and Adobe Photoshop on your Windows PC, so too should you be able to use Claude and Muse and Dot, together on your computer. They should be able to talk to each other. You should be able to script them. You should be able to use them from whatever interface(s) you want. Personal Computing should work on your literal laptop, or on any computer--that abandoned desktop in your closet, a Raspberry Pi, your old Xbox, and yes, even a computer that you're renting from a cloud provider. And you should be able to instantly migrate between any of these providers (e.g, if someone tries to jack up the price on you!)
One of the beautiful things about Personal Computing 2.0 is that, as much as these companies might not want it to be so, the genie is out of the bottle: open-weight AI models are nearly good enough that a model running on your laptop can use even their closed agents for you (by controlling your browser or creating a program to run on your phone). Not only does this enable the seamless migration from one computer to another, but it puts these "closed AI platform agents" squarely where they should be: in competition with one another. We want to avoid the platform-effects and other anticompetitive dynamics that contributed to the decline of the web, which means forcing these new AI companies to compete on cost, performance, and reliability--not on finding ingenious new ways of locking us in and "generating shareholder value" by extracting it from us.
Personal Computing 2.0 is naturally a more privacy-preserving way of relating to computing. The privacy concerns of the past two decades are largely caused by the shift in where our data lived. There were no real privacy concerns in Personal Computing 1.0 because it was obvious where your data lived: on your computer. You obviously controlled that. And by taking advantage of advances in encryption and cloud storage, Personal Computing 2.0 returns us to a world where the data is fully under your control and effectively solves most of the privacy concerns that have plagued us for the past decade.
But it's not just about privacy--bringing all of your data together allows for incredible new capabilities that simply are not possible any other way.
The olds among us will remember when Google Desktop existed (and was useful). For a brief moment in time, you could literally find anything, instantly. It was an incredible feeling, and an incredible product. The only reason it died was that--as the Internet increasingly became the center of gravity for our data--it just made less sense. It ceased to have access to all of our data, which made it progressively less useful until it eventually died. Today, we struggle to find and use the data that is scattered across our dozens of SaaS subscriptions.
But Personal Computing 2.0 changes this.
When all of your data is yours--when it's fully under your control--it becomes possible to find things again. To breathe more easily. To feel more relaxed. To be less scattered. As this happens, we may start to wonder:
Personal Computing 2.0 is about discovering the answers to these questions together.
Personal Computing 2.0 is personal. We all have different answers to these questions. That's part of the point! Just as we customize our home decor and our physical work spaces, we want everyone to be free to create whole new, personalized interfaces and systems and programs.
Personal Computing 2.0 is better suited to making these new systems in two important ways.
First, we now have access to a kind of computing power that simply wasn't available back then. This is partly about faster processors and more memory, but it's also about serverless compute: the ability to get 1,000 computers working for you for one second, rather than having to wait 1,000 seconds for your computer to do that same work. This enables applications that otherwise have required a massive cloud server and some SaaS app, and which we no longer need. But the computing power we have now is also different in the kind of compute that we can do: AI empowers anyone to make a computer do what they want in a way that was previously only accessible to professional computer engineers. AI unlocks a mode of interaction with our computers that was previously in the realm of science fiction: being able to talk to your computer, and have it understand what you actually meant.
But secondly, and more importantly, Personal Computing 2.0 is also about other people and our relationships.
It's about creating and sharing these new ways of computing and relating and exploring information. It's about building and dreaming and collaborating together. It's about re-thinking the past 20 years of computing and asking: what if our software worked exactly like we wanted it to?
One of my fundamental beliefs is in the innate goodness of most people. I believe that when people are healthy and happy and left to their own devices, they mostly try to help others. Most of us don't take joy from the suffering of others, nor do we care about accumulating infinite power. We mostly want to love and be loved, to find joy where we can, to offer kindness and compassion, and to leave the world a little better than we found it.
Personal Computing 2.0 is about using our data and our computers to seize back some of the power that Big Tech has acquired--and squandered--for the better part of the last two decades.
It's our data, our knowledge, and our power anyway. And there's nothing stopping us--let's start the 2nd Personal Computing revolution and build the world we want to see.