Artificial Intelligence; Facts vs Fiction

Artificial Intelligence (AI) is dominating the advertising industry, we’ve turned to one of our AI experts Silvia Sparry, Operations MD to help dispel the AI myth and uncover its reality

Artificial intelligence: A crash course

Discussions of AI have created a certain amount of unease, generating a fear that AI will evolve to take over the human race.  The likes of Elon Musk and Professor Stephen Hawking have suggested that there is a real possibility that super intelligent AI could become smarter than us, their human creators, likening this evolution to that of ours and our ape-like ancestors. In his “Ask Me Anything” session on Reddit, Hawking said “AI is likely to be either the best or worst thing ever to happen to humanity, so there’s huge value in getting it right.”[1]

In his “Ask Me Anything” session on Reddit, Hawking said “AI is likely to be either the best or worst thing ever to happen to humanity, so there’s huge value in getting it right.”[1] 

However, not everyone has the same definition of what AI really is.  To the general user the mention of AI will lead them to think of Hollywood robots, retina scanning targeting or the cool, ‘futuristic’ ideas like driverless cars.  The true definition of artificial intelligence is as the Oxford Dictionary defines it: ‘the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making and translation between languages.’[2]

The true definition of artificial intelligence is as the Oxford Dictionary defines it: ‘the theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making and translation between languages.’[2]

Silicon Valley giants such as Google, Microsoft and Amazon are seen to have used AI to develop, grow and improve their business.  Projects such as Google AI and Microsoft Brainwave have been developed to “wow” their audience but are in fact a result of machine-learning, a common buzzword that is often used interchangeably with AI but actually has a completely separate meaning.

While AI is the broader concept or skill of machines being able to carry out tasks in a way that we consider as ‘smart,’ machine learning is the technique where through the consumption of data machines are able to learn for themselves.  It is machine learning that is most often referred to in today’s AI conversations.

Machine Learning: The everyday function

With a wide variety of digital functions that we use in our daily lives it can be difficult to see how frequently we use AI, powered by machine learning. The reality is that it’s used in pretty much everything; the spam filter in our emails, image recognition on social media, product recommendation on shopping websites, recommended music lists on Spotify, voice recognition and smart ‘personal assistants’ such as Siri or Alexa.

Machine learning is simply the clever application of algorithms to data to make predictions. Simply put, it effectively turns input A into output B. Feedback loops can further refine results: Someone (or something) reviews the original predictions to decide if they were accurate, thereby increasing AI intelligence for future predictions.

What makes AI so attractive for advertisers?

The impact of AI systems comes down to the way they are used.  By using them to supplement human intelligence, every day or mundane and repetitive tasks will be eliminated or made easier, freeing people up to do the tasks AI cannot do. Computers can also perform these tasks at a speed and frequency than humans could never compete with. For example, AI has become hugely relied upon in the banking sector.  By continuously scanning location and other detailed data inputs for suspicious patterns, AI can help banks and credit issuers identify fraudulent behaviour while it is happening allowing for the issue to be addressed earlier.

At Xaxis, our artificial intelligence platform, Copilot, has been designed to optimise campaigns and support traders in their quest for improving performance.  AI is capable of discovering patterns and outliers in large data sets while humans are able to apply broader context and draw additional conclusions from those patterns. The Copilot platform enables our Campaign Managers to have visibility into what decisions the AI is making and allows them to adapt this where necessary to deliver on custom advertiser KPIs. They can also design their own optimisation algorithms, which is useful when they have found an optimal optimisation approach for a campaign and then give the task of executing this optimisation logic in the DSP to Copilot. In this way optimisation approaches developed by our teams of programmatic experts and Data Scientists can be translated into customised algorithms designed to drive advertiser business goals. Through this combination of human experience and AI we can discover unique pockets of performance much more accurately and faster than using either approach in isolation.

We therefore believe that artificial intelligence is best applied to improving and supplementing, not replacing human intelligence. It’s time to dispel the myth that artificial intelligence will replace human kind and educate users to understand its unique capabilities that can enhance our world.

[1] Stephen Hawking: Artificial intelligence could wipe out humanity when it gets too clever (Independent, 2015)

[2] Definition of artificial intelligence (oxforddictionaries.com)