President’s address
Artificial Intelligence (AI) is not just another buzzword in our business dictionary. It is not just the claim of companies that want to get more exposure or increased funding. Whilst it certainly has all the elements of a buzzword, including a value bubble, it is unarguably a game-changer for our profession; a game changer that opens up a whole bunch of opportunities for insights professionals.
Today, AI plays a fundamental role in easing the day-job of researchers while offering a totally new set of capabilities. Capabilities that can be divided into two big segments: those that boost the collection and treatment of consumers’ data and those that foster the interpretation and outcome of the data.
Examples of the first could be processing and coding massive datasets from surveys or from social media listening, transcribing and translating open-ended responses in multiple languages, converting speech to text, recognising content in image and video, biometric data processing and facial and eye recognition among many others.
Segmentation, predictive models, attribution analytics or a digital path to purchase analysis could be included in the second. All these solutions powered by AI allow researchers to get into new fields of consumer understanding and deliver an unprecedented value.
However
Black boxes are the flip side of AI. Let me share some data: 37% of the worldwide internet traffic on mobile is made on YouTube (Sandvine). YouTube’s AI is responsible for 70% of views by computing recommendation. This means that one AI model is responsible for 26% of all internet traffic on mobile. A single AI model, unknown to its users, has a huge impact on what people view and, therefore, an immense influence on people’s behaviour. What a responsibility!
Users have a very pragmatic approach to AI. As long as it delivers what we think is good, we have total trust in the algorithms with no questions asked. Is this blind trust in algorithms the right approach? Technology is not neutral. It mirrors the domain knowledge, the respect for ethical codes and biases of its developer. And it is precisely on this critical stage that the 80+ years legacy of the insights profession can and must play its role.
There is a superior opportunity for insights professionals. This is to lead the implementation of AI and machine learning in the data analytics democratisation process to where we are heading – a democratisation of data and insights tools that will increasingly be used by non- experts. Every day more data and insights will be used by professionals who lack key knowledge about what they are dealing with. Professionals with no research skills and no understanding of how the insights provided will have been obtained.
As we play the video suggested to us by YouTube, they will play the insights generation button. Isn’t this a good enough reason for insights professionals to embrace AI and lead the transformation of the industry?
