User friendly tools may put the power of artificial intelligence (AI) into people’s hands, but without the right specialists to light the way, there’s no guarantee of success.
Recently Google’s AI AlphaGo Zero proved it didn’t need to watch people actually play the ancient Chinese game of Go in order to learn how to play it.
Timeline for AI
- October 12, 2017
AlphaGo Zero taught itself simply by playing the game — it would appear AI is better at playing Go than people but that AlphaGo Zero doesn’t need people to master the game.
The same holds true for AI in business, where advancements within the AI realm of machine learning allow computers to learn without being explicitly programmed.
These algorithms are so pervasive that we take them for granted: When your phone shows you the best route to work; when your bank notifies you of a potential fraudulent charge; when you TV recommends that you spend the next ten hours binge watching Game of Thrones — these were all driven by self-learning simulations that use historic data to predict future outcomes.
It is no surprise then that business users are bullish on machine learning.
Almost 40 percent of those surveyed by Verdict and GlobalData have already invested in machine learning technology, beating out mega trends like bots and self driving cars.
Businesses have built up an alarming familiarity with AI and an expectation of simplicity, where any programmer can, with a few lines of code, create and deploy a working predictive model.
With the likes of Microsoft Internet of Things (IoT) Hub and Azure Machine Learning a developer can quickly stand up an IoT app that can predict the weather based on collected temperature and humidity data – no muss, no fuss, and only a few select lines of database code.
Such tools are amazingly accessible, allowing almost all businesses who want them to make use of AI.
But, in doing so, these products also paint a picture of AI as just another tool in the developer’s toolbox, something we can put into action quickly and with immediate effect.
And that is a very dangerous assumption.
How do we know the supportive weather data is correct or appropriate to our needs? How do we know if our model will remain accurate over time? How do we even know if we’re correctly interpreting the predictive outcome?
We should, therefore, view the current machine learning gold rush as a warning – a reminder that the application of AI in predicting business outcomes is an endless cycle where collaborators seek to answer specific questions.
More importantly, it reminds us that AI is a team sport, requiring the cooperation of many specialists — not a single, PhD-equipped data scientist, mind you — but instead a small, diverse team of pragmatic experts capable of reliably employing tools like machine learning algorithms.
Some are hiring consultants that specialise in AI technologies like data engineering, machine learning and so-called deep learning, optimisation and prescriptive analytics, business and vertical markets, and communications (someone who can tell compelling, accurate stories) to make better use of AI.
Machine learning algorithms are incredibly powerful and the likes of Google, Microsoft, Amazon, and Salesforce realise know this — which is why they are investing so heavily in putting them to use.
But, those algorithms alone are no guarantee of value on their own without people.
Whether you’re predicting the weather or optimising a delivery route, AI lives or dies according to the humans willing to not only nurture but also hold AI accountable.