AI-powered super-robots
The possibility of AI-powered super-robots taking over the world has appeared in fiction in lots of fun ways over the past 50 years or so.
Black Mirror, Westworld and classics like The Matrix all depict worlds in which data- & AI-driven digital beings call the shots a little more than we’d like them to.
It’s easy to dismiss the above examples. They’re fun, sure, but ultimately... a little far-fetched. However, concerns about AI replacing humans at work are closer to home.
At Datasparq, we believe that AI will augment human intelligence, not replace it. AI will enable humans to make better decisions, get more done and free up their time to work on the stuff that really matters.
“Great servant, terrible master”
The bottom line is that AI makes a great servant, but a terrible master. What do I mean by this?
AI is a great solution for solving complex problems but without human experts guiding and directing it, it won’t be able to do much at all. In a business context, it’s the experts on the ground that need to work with the technology to unlock its value.
For example, when businesses use AI to power their dynamic pricing strategies, revenue managers and pricing analysts work alongside AI to turn their deep analyses into real prices that work for the customer and the business. Computers are adept at solving mathematical problems far beyond the grasp of any human—but need human experts to guide them both in solving the right problems—and in knowing where to start.
An ML model needs to be fed the kind of data we think is likely to have an impact on the outcome we’re predicting. For example, as someone who’s worked in TV, I know that weather impacts viewing patterns. Therefore if I were to build an ML model to predict viewing figures for a particular channel next Wednesday afternoon in London, I would add weather data to the training set.
AI as a product
To realise value from AI—as I mentioned earlier—you need a team that knows how to train it, use it and apply it to real-world situations. Without that, it doesn’t matter how sophisticated your tech is, its value won’t make it into the real world.
As the Product Director at Datasparq, applying product thinking to data science and engineering projects is a core part of my job. The ultimate goal is to ensure that people can actually use and get value from AI. For that to happen, people and processes must be at the core of the project.
To that end, at Datasparq we don’t treat AI projects as pure technology projects. Technology is one part of the puzzle, but effective AI solutions come from multidisciplinary delivery teams (design, product, data science, engineering), working closely with end users, data experts and IT. This way we can ensure that our ML products are more than just great data science. They are usable, resilient and solve real problems for the business.
Ultimately, treating AI as a product is essential to unlocking the value trapped in the treasure troves of data in your GCP, Azure and AWS servers.
