There is little argument amongst technology industry’s analysts that the rate of app creation and device proliferation is set to grow at unprecedented rates. What are considered to be conservative estimates has the number of deployed devices exceeding 200 billion with over 25 million supporting apps in place by 2025. The volume of data associated with that number of apps is even more difficult to predict. Needless to say it’s likely to exceed trillions of gigabytes in size.
The challenge we face as an industry is how do we create and support that number of apps to service the demand? The traditional development methodologies and outsourcing business models seem to be inadequate when considering the task ahead. Furthermore, the pace of app change compounds the problem. As businesses embrace a continuous innovation model to remain relevant in an ever increasing global market, so too must the apps adapt to support that innovation. By all measures, the demand upon the resources available to create and modify these apps is insufficient to meet the market need.
The traditional development approach has served us well for decades but we are now at a point of inflection. Current trends are seeing development tools and platforms beginning to empower a larger demographic of potential user. The term ‘Citizen Developer’ has appeared to imply that the development of apps is beginning to embrace a larger audience other than career software developers. The creation of apps is trending away from traditional app vendors and IT departments and moving more toward the business line where the domain expertise and competitive innovation originate.
We’re currently witnessing an evolution of sorts. Several years ago, there was a lot of discussion around big data and investment made into big data ventures. That interest has now shifted to machine learning and will evolve again as the industry looks to the role of artificial intelligence. The three are not disconnected. Learning requires input which, in our scenario, is the data that we gather as a function of the apps we use. Artificial intelligence relies on that learning to apply action to that we deem relevant to the required task. That is, we need data to learn and learning to drive behavior. The industry today is still largely looking at machine learning without taking the next step to apply that learning to an actionable outcome in an automated or semi-automated manner.
At Crowd Machine, we refer to actionable outcomes as a pattern of behavior. We allow the definition of patterns as suites of related and non-related behaviors. These behaviors can be rapidly created, modified and reused to meet the exact needs of the business in real time. The Crowd Machine Design Studio has been purpose built to empower developers and non-developers alike to create apps by defining patterns of behavior that combine to form the desired outcome.
Underpinning the process of creating apps using Crowd Machine, are the principles of machine learning. Crowd Machine applies machine learning to the development process to accelerate app creation and reduce the risk of erroneous outcomes. This has the effect of not only getting an app to market significantly faster than previously thought possible, but also ensures a high quality of app delivery. The application of machine learning doesn’t end there. All apps created using Crowd Machine are themselves machine learning solutions. These apps have the ability to identify behaviors in both their data outcomes as well as how they’re used. The intersection between outcomes and how those outcomes were achieved provides insight into how the app itself can be modified to greater effect.
At Crowd Machine we started the journey toward the application of AI to app development at the beginning of our technology’s development. The patterns of behavior that make up an app are essentially blueprints that control the way in which our Crowd Virtual Machine (CVM) functions. Think of it as a set of instructions that cause the CVM to morph into the app that the instructions define.
Patterns themselves are data and as such can be altered in real time. What this implies is that the application of artificial intelligence (AI) to the learning process can dynamically modify an application at runtime by altering the pattern definition. This in turns alters the CVM function. We’re essentially creating self-evolving apps that change their form and function at runtime based upon learning from their environment.
At Crowd Machine, we believe we’re on the cusp of creating a software paradigm where software wil self-evolve from learning. In theory, the result will be the creation of software that will constantly change to meet the dynamics of its target market. Developers will eventually become nurturers of the softwares evolution as opossed to writing code. We’ve already started down this path with Crowd Machine today and there’s much more to come.