Being a data scientist has been a hot job for several years. Concurrently, we have more and more questions about ethics and who the industry really develops value for. If you are a data scientist and want to create state of the art things, you probably must work for one of the giant data companies and accept their models to utilize data. Could this change soon?
We still remember the days, when you had to work directly or indirectly with a mobile carrier or Nokia, if you wanted to make an application for a phone. It also had a very small likelihood that your app ever got to consumers and basically you did what some business folk had decided people wanted. Then came Apple’s App Store and overnight everyone in garages or bedrooms around the globe were able to make and publish mobile apps. It became a consumer market.
We also know very well what it’s like to work for a bank and create finance services. You are a part of a huge machine and only work on some small components. No wonder, when blockchain opened the market to finance services, it activated millions of people to develop things. Of course, it hit some hype too, which often happens, and we are still in the early days of how distributed ledgers will change many services.
How could the same happen to data scientists and AI developers? Or is it so that Google, Facebook, Amazon, the NSA and some others dominate the data market so overwhelmingly that no one can challenge them? Or at least so that individual developers or small companies cannot ever compete with them?
Personal control and ownership of data are becoming very important. Privacy regulations, like GDPR in the EU and CCPA in California, are only one part of that. More and more companies are emerging to develop solutions for personal data control and also many big companies are starting to see the benefits from the new model. More companies could better compete against the data giants and they could also decrease their own risks and liabilities, if consumers could keep their own data.
There are also technology needs to make more distributed data and AI solutions. For example, many personal assistant-type services require availability, latency and security where it would be better to have local data that is utilized in analytics and AI. We would move from very centralized massive big data cloud services to distributed data in local devices and consumer’s own repositories.
All this will also change the data application market and how to generate business with them. It opens a market to new actors and also independent developers to offer their applications direct to consumers. When consumers have their own data, they are able to utilize many new services and applications in their daily lives. It is similar to over 10 years ago in the mobile application market.
Of course, this needs many components in the ecosystem until it really works properly. We need a framework to develop these applications, an active developer community and enough parties to orchestrate the ecosystem and services. We already see a lot of development in this area, so there are probably not so many missing components anymore.
Whereas data scientist work has been to develop algorithms, make more or less advanced data mining or develop some ad or sales targeting services, this new development could change job descriptions significantly. Those things are all still needed, but there will be more opportunities to innovate totally new services on data, make new business models for consumer data apps and start to offer AI apps directly to consumers too.
We are approaching a disruption point in data services. It won’t only be about privacy, consumer control or distributed data and AI, but it will also introduce a significant change in how data services and applications are developed. It opens business opportunities to new companies and developers. This is development that has also happened in other software areas, from centralized business management driven systems to more independent services, an open market and individual developers. It gives more freedom to consumers, how they want to use data. And for developers, what kind of data applications they really want to develop and bring to market.