Articles > AI for Business - Your 5 AI reads of the week #3

3 months ago

Welcome back to your weekly dose of AI for Business news. By now you know the drill, first we highlight the figure of the week and then present to you 5 articles that show the true potential of AI in different types of business fields. This week, you will learn about: the mindset that goes with AI, some basics about Machine Learning and Emotion AI, and finally some AI applications by McDonalds and ... a Finnish prison.



83%: the percentage of AI adopters that have already achieved substantial (30%) or moderate (53%) economic benefits. This was one of the conclusions of a Deloitte survey of 250 companies who have invested in Artificial Intelligence. Hence, only 17% of respondents said their company had not been able to realize an economic benefit so far. Furthermore, the survey suggests that there is a correlation between the experience with AI and its benefits, the more AI projects the higher the percentage of projects that realized economic benefits. This should motivate you as a company to not think of AI as a one-time project, but invest in different small proof of concepts and find out which implementation of AI has the biggest impact on your business. To put a number to this all, the survey found that companies who had deployed 11 or more AI projects, had a 92% chance of yielding economic benefits.


CIO’s should also adjust their mindset accordingly for AI projects compared to incremental projects. Gartner found that 37% of organizations are still looking to define their AI strategy and 35% have difficulty to identify use cases. The latter companies should take a look at our AI reads, plenty of use cases to choose from ;) But indeed for any IT projects, the decision to invest often arises from a company’s specific needs and thus a clear business case is crucial to understand where an implementation could be possible. AI is still a new tech, therefore it is also harder to predict the costs and return on investments of such projects. But it is possible. Another barrier for the CIO’s is that AI projects call for a diverse set of skills and thus talent acquisition becomes crucial but also another pain to deal with. There's an increasing trend of partnerships between corporates and AI start-ups to fill in this gap of talents.

If you are someone who does not know a lot about Machine Learning but has basic knowledge in mathematics, this article will help a lot in understanding ML without the coding part of it. This article is great for the non-engineers because by understanding ML it will be much easier to find use cases for your business. It is always easier to find implementations when you understand the idea/concept behind different solutions such as AI or ML.

Emotional AI is a subset of AI that measures, understands, simulates and reacts to human emotions. Humans might be better at reading emotions and the environment to understand how to feel and react, but machines are the best at analyzing huge amounts of data. So, we’re slowly starting to see the first machines that can understand, by the tone of your voice, whether you are stressed or happy. According to this article, the following industries have already benefited from emotion AI: call centres, mental health monitoring, automotive and assistive services.


The Nordic countries have always been known to be very progressive, however, this still came as a surprise. A Finnish company could not find enough people to label data for Finnish websites, so they found help from the local prisons. The prisons were excited because this type of job only required a laptop, there is no risk of violence in comparison to jobs where one would need tools, which could be transformed into weapons. The partnership has been successful because, within 2 local prisons, around 100 prisoners are working for this Finnish company. The article raises questions whether this type of labour could be possible in less progressive countries such as The United States. The opinions are mixed, either it is seen as economically exploiting prisoners for a very low wage, or as a helping tool to rehabilitate because you do not come in contact with possible weapons. What is your opinion? Is it ethical to use prisoners to label data in order to train an AI? Let us know via Social Media!

Last week, McDonald's closed its largest acquisition in the last 20 years: an AI company called Dynamic Yields. The goal of this acquisition is to have better variations on their electronic menu boards displaying items depending on factors such as weather, holidays and regional preferences. In other words: personalized customer journey for increased profits This article shows another big player who has realized the potential of partnering up with a start-up that is focused on Artificial Intelligence! What are you waiting for ;)

That’s the end of the third edition already! I hope you learned something new and see the many implementations AI can have in different industries!

alt text