The insightWhen you start out with AI in your organisation, you most probably have a clear picture of what you would like to achieve. You know the processes to optimise, the data to analyse etc. However, you still need to translate these business objectives into concrete AI tasks. The following heatmap shows which AI tasks and paradigms are likely to be relevant for your industry:
Relevance of AI tasks by industry
Reading the chartThe heatmap was built based on an NLP analysis of ca. 50k AI-related articles in the timespan 2020 - 2021. The rows represent the AI tasks and paradigms, whereas the columns represent the industries. The higher the score in a cell, the more relevant the respective AI task is for the considered industry. Visually, higher scores have darker colours.
Interpreting the chartMany of the values in this chart are quite intuitive. For the higher values, you would expect that Fraud Detection is highly in demand in the Finance industry, that Computer Vision is essential for assisted or autonomous driving etc. On the lower end, we are not surprised by the weak associations between AGI (Artificial General Intelligence) and most of the industries - because, well, AGI is general.
However, upon closer look, some of the findings are quite intriguing - let's consider three of them:
- AI is making a dent in Agriculture. In particular, Cognitive Computing - a technology classified by Gartner as "obsolete before reaching the plateau of productivity" - is showing strong momentum. Being especially apt at modelling organic thought processes, Cognitive Computing is supporting the evolution of agriculture towards more sustainable practices (cf. for example the COGNAC lighthouse project by Fraunhofer).
- Finance and Healthcare are the industries that show the strongest intersection with Artificial Intelligence. At first sight, this might appear counterintuitive since these two industries are also heavily burdened by regulation. However, with AI thriving past regulatory restrictions and towards more transparent, responsible algorithms, we can see that it also takes foot in tightly regulated industries. Keep in mind that a big part of the AI buzz is happening in the innovation hubs - not your traditional bank or hospital, but the visionary fintech startup or R&D department of a pharma company.
- Explainable AI is one of the important AI trends this year. We can see from the chart that it has a high importance for Agriculture, Finance and Healthcare. But - what about Automotive and, more specifically, Autonomous Driving? Is the ability of an Autonomous Driving system to explain and justify its decisions not crucial to forge trust among consumers? Looking at the chain of heavy and partially deadly road accidents caused by AI in the past two years, it looks like Autonomous Driving still has to mature not only in the quality, but also the transparency and explainability of its decisions.
What do you think when you observe the chart - are there any other datapoints that surprise you? Feel free to post in the comments so we can figure out the AI puzzle together!