August 22, 2018
The new research from Technavio has revealed that the global enterprise AI market will grow by over 5 billion US Dollars by the year 2022.
This research also provides deeper insights into the contributions of each industry sector thanks to this growth, with the main players being Telecommunications, Advertising, consumer products and services, and the Healthcare Industry.
With the pace that data analytics and other technologies are growing, this amount is not surprising. In fact, the part that astonishes readers the most is that Chatbot Artificial Intelligence is among the top drivers for this drastic growth. Which is unexpected because, until now, chatbots have failed to meet the upraised expectations that analysts have set. However, with the advent of Chatbot AI, unveiled in Google I/O 2018, things appear to be picking up the pace.
This could be because these new and more advanced chatbots will not just be a cheaper alternative for the human counterparts in the field of customer care. Rather they will be a more fitting and able provider of customer services, with better sympathising abilities than most humans. It’s not like these chatbots will be able to actually empathise with the customer, it’s just that they won’t have their own emotions to cloud their judgement, and as a result, they will be better able to solve customer issues and queries, by taking assistance from the background data and historical records.
This application and its foreseeable effects also bring more light to the less explored applications of AI, such as market research. Only recently, an AI expert and CEO of an Artificial Intelligence startup, Andrew Konya published a detailed report on how the technology can bring revolutionary changes in the area of market research on people.
Sampling, and its less effective application in deriving the population data, has always remained the major pain point in statistics. After all, any small group, no matter how carefully chosen, can never fully represent a larger group of people. As Andrew pointed out, “People are more than data points”. The most common alternative for this is extensive surveys, even though these have proved to be even less effective, mainly due to the bare minimum and closed-ended questions. They can always tell you the precise ‘what’ part of the question being put to the audience, but never the ‘why’, which more often than not holds more value.
Konya suggests that both these problems can be solved with the proper application of AI in market research. In one case, he clarifies how AI and its ever-growing prowess of analysing the huge amount of complex data, provides the quantitative advantage in the research, negating the setbacks of surveys. On the other hand, the algorithms of deep learning trained on the particular niche, can also provide more in-depth insights on the sample data, adding the qualitative advantage as well in the bucket of AI-based market research.
To prove his point, he also illustrated the application through a real-life case from the city of Cornwall. For many years, the city’s residents were living in isolation for elongated time periods. They would spend months without speaking to any of their friends or acquaintances. This was leading the city deeper into the pits of loneliness, compelling the local community to come up with a solution.
Not long ago, the council launched a program, in partnership with Deloitte, to collect active data from 200 participants. Active data is different from the ordinary data in terms of how it is collected. To gather active data from a person, the AI has to actively engage with over 200 people, asking more contextual questions on the basis of their previous responses. Data collected with such measures proved to provide more actionable insights. In this case, the community was able to learn what existing programs were actually effective, and how the local communities can better intervene to bring positive changes in the social care facilities of the city.
This is just one use case out of many where active data based research has produced more reliable outputs quicker than before. It is therefore clear that we can realistically look forward to progress in AI.