Understanding Actor-Centric Computing
In the first article of the series, I introduced the idea of a new paradigm that move away from our familiar discrete computing devices understanding to move to an actor-centric one. In this article, I want to explore a little more what it means and the benefits we get from this new way of thinking about software.
Imagine a world where there’s no computer to buy, no new cool version of your favorite phone, no need to exploit Earth’s resources to create another phone that will finish in your graveyard of old phones next year. For me, it’s innevitable that we need to move away from our computing as a product thinking, which put environmental pressure to grow corporation profits, and start thinking about humans and other actors as the fundamental building block of a new computing era.
Do you need to buy your own power plant to get electricity? No, you don’t. You find a provider and subscribe to a service that will power your home for a certain monthly price, based on your consumption needs. Why don’t we think the same way for computing? We already do it for our phones, even getting the phone from our provider, so why couldn’t we think of a world were you subscribe to computing resources and use them just by plugin in a socket in your home? Right now, our mindset is based on a “device first” understanding of computing, inherited from the early age of information science.
Imagine now that you are in your home and you have powerful implants in your brain that are able to render your thoughts as actions or visually render the information you receive directly in your eyes (or brain). In today’s computing, you would have to become a computer, with installed software and bi-annual upgrades! This would be similar to buying a VR headset today, the device comes first, humans are external users.
Now, let’s assume that we reset our computing paradigm and base everything on the concept of actors, no physical devices. An actor is an entity, with a strong cryptographic identity, that can trigger actions and serve as logical content container to manage its own private content. An actor could be a human, an object, an organization, a system or an AI agent. By default, all actors are alone, isolated, but they have the power to establish explicit content subscriptions between them, similar to following someone on your prefered social network. So, fundamentally, our new computing paradigm would be a mesh of interrelated actors, reactively and securely connected together, sharing part of their content between each other.
These logical actors don’t have their own computing capacity. They need to subscribe to enough computing capacity to support their needs. For example, a human actor might need a runtime environment to execute logic, storage to store media, documents or other structure content and some kind of human-machine-interface to be able to interact with their content. A system actor might subscribe to more runtimes, storage and maybe some specialized services but no need for a HMI renderer.
As you can see, in this paradigm, computing devices don’t dictate the topology of any information system. We can’t build information systems or release them, as they are dynamically composed by actors based on their needs and interests. We don’t have to create complex functional requirements models and we don’t need to create snapshots and develop a product for months of effort. Each actor is unique, responsible for subscribing to other actors based on its specific needs. For example, your role is to approve invoices over 25K$? You subscribe to your company invoices stream, apply a filter for invoices over 25K$ and create a simple list of invoices to approve in your personal space. This list is private and only visible by you. You just need a simple form to review the invoice, change the status as needed and publish your updated invoice in your own approved_invoices stream. Other actors might subscribe to your approved invoices and perform some tasks, as part of large actor-driven chain reaction. Ultimately, the invoice will trigger a payment, etc. This can all be automated and modeled organically, without the need to predefine complex business processes or inter-organization collaboration.
Without digging any deeper for now, it’s easy to see how changing the perspective on software, from hardware to actors, can open our mind to new solutions, more aligned with the modern problems we need to solve. There are much more to explore, better understand computing providers, new business models for hardware manufacturers, actor relationships, how can we help humans be responsible in managing their own information systems and much more. We’ll explore these subjects in details in this series.
