Can AI read our emotions?

9

September

2022

No ratings yet.

As new technologies become ever more widely accessible, the business are looking for ways to further automate our lives. The idea is simple (and quite convincing!)- why should we spend time on operations that AI can be trained to perform? In a lot of cases the AI/ machine learning solutions are even more effective, as no human error is at play. One of such technologies, which are revolutionizing some industries is emotion recognition (Sydorenko, 2021). Companies like start-up Emotient have produced software which, they claim, can identify people’s emotions (Crawford, 2021).

The emotion recognition tools are now being applied to quite a large array of activities: they can be used to analyze you during your job interview and airports used them to identify potentially dangerous people. Let’s take HireVue recruiting company for example- in 2014 it has released a software that analyzes job applicants’ faces and voice tones. They then compare it with the best workers in the particular company in hope that these attributes can identify employees that will bring the most value into the company (Crawford, 2021).

We could argue that systems like this are wonderful examples of how much our society can benefit from the technology. There is just one issue: all of these emotion recognition systems are based on a theory that all humans experience a few universal emotions and that they are natural, innate and not dependent on the culture we grew up in (Sydorenko, 2021). This theory hasn’t been in fact yet proven by any reliable and detailed research (Gifford, 2020). Most believe that Paul Ekman, American psychologist, has proven this theory, but the scientific circle (eg.Margaret Mead, American cultural antrophologist) has been sceptic about his research methods and assumptions (Crawford, 2021).

For me it seems very irresponsible to create tools that can, presumably, read people’s emotions and ambitions, while scientists haven’t even been able to prove that emotions can be read from human expressions or their voice. It assumes all the emotions are also expressed in the same way by all of us, which is hard to believe. In this shape, the emotion recognition technologies are often putting people in a disadvantaged position (rejecting from a job/ identifying as a dangerous individual) without any reliable reason.

Do you think it is a dangerous phenomenon? Or rather an advancement towards a more technologically operated world?

References:

Sydorenko, I. (2021, August 25). AI in Emotion Recognition: Does it work?. Label your Data https://labelyourdata.com/articles/ai-emotion-recognition 

Crawford, K. (2021). Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence. Yale University Press. 

Gifford, C. (2020, June 15). The problem with emotion- detection technology. The New Economy. https://www.theneweconomy.com/technology/the-problem-with-emotion-detection-technology

Please rate this

Sentiment analysis: How a computer knows what you’re feeling

5

October

2021

No ratings yet.

One of the most distinguished features of humanity, is being able to read someone else’s mood. People estimate other peoples mood and emotional state based on their verbal and non-verbal communication. Humans learn this skill from a very early age on and are able to distinguish the signs even for different people. For example, we all know when our best friend is feeling down, even when they’re trying to hide it for others. Maybe by being louder than usual, or they might be more quiet. But you know that they are troubled. Being able to distinguish this for different people is due to experience and how close you are to the other person. The years of friendship make it easier to read her mood and know when something is wrong.

Sentiment analysis

But what if I tell you a computer can also do this? And they don’t have to ‘be your friend’ for years and years. Using various techniques such as Natural Language Processing (NLP), the computer is able to recognise words and sentences as emotions. Words like ‘stress’ and ‘feeling alone’ are registered as negative, whilst ‘glad’ and ‘exited’ as marked as positive. This is called a sentiment analysis. Some sentiment analysis even link certain combinations of words to feelings, such as depressed, sad, cheerful or hopeful.

Possibilities

But what are the possible use cases of the sentiment analysis? And what are the challenges? Sentiment analysis is for example already being used in healthcare. Based on the unstructured notes a psychologist takes during sessions with their client, the sentiment analysis can track the clients mood over time. In this case, it includes possible diagnoses and ‘trigger words’ such as suicidal. This allows the psychologist to have one overview of the client’s emotional and mental state over time.

But what if this is taken one step further. What if this is done based on social media. Think about it: Your stories, chat messages, posts, web usage, emoticon usage, everything is documented and examined on your emotional state.

Challenges

One of the most important questions that arises from this, is regarding to privacy. Who is allowed when to measure your mental state based on your social media usage and your google searches? This is very sensitive (health)data, not stated by a doctor, but estimated by a computer. Another challenge is interpreting the words correctly, for example: ‘good’ is a positive word but ‘really not good’ is a negative combination of words. The complexity of languages can make it hard for a computer to interpret it correctly. However, Artificial Intelligence (AI) is often used for this so that the estimations become more accurate with each analysis.

Sources

Abualigah, L., Alfar, H., Shehab, M., & Hussein, A. (2019). Sentiment Analysis in Healthcare: A brief Review. In M. Abd Elaziz, M. Al-quaness, A. Ewees, & A. Dahou, Recent Advances in NLP: The Case of Arabic Language (pp. 129-141). Switzerland: Springer, Cham. doi:10.1007/978-3-030-34614-0_7

Denecke, K., & Deng, Y. (2015, March 25). Sentiment analysis in medical settings: New opportunities and challenges. Artificial Intelligence in Medicine, pp. 17-27. doi:10.1016/j.artmed.2015.03.006

Please rate this

How Digitization is Destroying Your Emotions

26

September

2017

No ratings yet.  

Remember that book you loved reading? How you felt when you opened it and started a new chapter? Or that one professor that used to talk so interestingly you could not not pay attention to him?
When you imagined these things, you probably thought of a physical book and a professor in a physical lecture room, not about an e-book or an online course, right?

Products like books – aside from required lecture books maybe – are usually categorized as hedonic or experiential goods; people purchase them for the pleasure they get from the product (Chen & Granitz, 2010). Also, people tend to attach more emotional value to physical experiential goods than the digitized versions of those products (Waheed, Kaur, Ain & Sanni, 2014). Nowadays  digitized versions of a wide variety of products exist, called digital information goods (Goh & Bockstedt, 2013). To give an example: sure, you can purchase a Beatles album straight from iTunes as a birthday present for your Uncle, but wouldn’t he be so much happier if you would give it to him on vinyl – given your cool uncle has a record player, of course.

Now, the question here is: How can digitized goods provide just as much emotional value (if not more) as physical products?

Note that the question is not whether hedonic products should be digitized at all, because, of course, it is way more efficient to carry around 500 songs on your phone, instead of carrying them around on CD’s. So, when we look at efficiency, digitization is a big help. Also, selling information goods like these can be very beneficial for companies. Creating that first product might be expensive and takes some time, but creating the second version is just a matter of making a copy and, consequently, takes very little time and resources. In other words, marginal costs of information goods are very low, meaning companies can enjoy a big profit marge (Brynjolfsson & Bakos, 1998).  However, wouldn’t it be nice to be able to be just as happy about digitized products as you are about physical ones? If this can be achieved, this could have major effects on companies still producing physical goods that can be digitized, think of DVD’s – which have already lost a huge chunk of market share because of Netflix, HBO, etc. –  postcards and even schools may be at risk somewhere in the unforeseeable future.

Efficiency of digital information goods. Source: https://theecoguide.org/books-vs-ebooks-protect-environment-simple-decision
Efficiency of digital information goods. Source: https://theecoguide.org/books-vs-ebooks-protect-environment-simple-decision

One thing that can be done is to offer more customer value in the case of digitized products. For example, if you purchase that Beatles album – or any other album – through iTunes you get free extra’s such as little video fragments of behind the scenes footage. In this way, consumers get more value for their money, which might increase their happiness.
Although there is no solution to this problem yet, companies selling digitized hedonic/experiential products should try to get consumer happiness to the same level as consumers get from physical products in one way or another.

So, next time you’re thinking of sending someone an e-card, remember that they’ll probably be happier to see one on their doormat.

 

 

Sources

  1. Brynjolfsson, E. & Bakos, Y. (1998). Bundling Information Goods: Pricing, Profits and Efficiency. Management Science, 45(12), 1613 – 1630.
  1. Chen, S. & Granitz, N. (2010). Adoption, rejection, or convergence: Consumer attitudes toward book digitization. Journal of Business Research, 65(8), 1219 – 1225.
  1. Goh, K. H. & Bockstedt, J. C. (2013). The Framing Effects of Multipart Pricing on Consumer Purchasing Behavior of Customized Information Good Bundles. Information Systems Research, 24(2), 334 – 351.
  1. Waheed, M., Kaur, K., Ain, M. & Sanni, S. A. (2014). Emotional attachment and multidimensional self-efficacy: extension of diffusion theory in the context of eBook reader. Behaviour & Information Technology, 34(12), 1147 – 1159.

Please rate this