Artificial Intelligence has a lot of positive impact on the world. Processes are being standardized and accelerated, the error rates are dropping, it operates 24×7 without interruption or breaks and has no downtime (etc.).
However, there are some downsides of Artificial Intelligence. Looking at the recruitment market, AI can cause some problems. While employers use AI to automatic scan resumes, the software can make false assumptions to discriminate the candidates. Automated hiring systems, such as resume scanners, are stopping an estimated 27 million people from finding full-time work (Fuller et al., 2021). The employers who are using these techniques to scan resumes are turning down possible candidates. According to a Harvard Business report, the scanning tools focus on things that are missing on the resume, instead of the value someone can bring into a company (Fuller et al., 2021).
Certain groups are more affected than others. Caregivers, veterans, immigrants, people with disabilities, prison leavers, and people whose spouse has relocated are the most affected groups, the report said. This can give them a huge disadvantage on other candidates, what can cause that these people can’t find a job.
Next to all the benefits AI is bringing us, we have to be aware of the danger and downsides of it. In my opinion, it is very useful and cost reducing to use AI driven technologies like resume scanners. However, we can not discriminate some groups because of their shortcomings. In this way, these groups wouldn’t take part of the society just like others. The technology should focus on the competences of candidates. Or, human recruiters could finally check the rejected resumes to see if someone can go through the process again.
Reference
Fuller, J., Raman, M., Sage-Gavin, E., Hines, K. (September 2021). Hidden Workers: Untapped Talent. Published by Harvard Business School Project on Managing the Future of Work and Accenture.
Hi Guus! Nice blog! I agree with you that AI has a huge potential but that it also has a lot of downsides. Next to discrimination there are a lot more bad things considering AI as: unjustified actions, biases, opacity, autonomy & privacy and moral responsibility. Do you agree that these cases all are problems of AI? (these problems are all mentioned in the lecture from Tamara today).
What do you think of the possible solution of the ethical guidelines from the EU? I think it is a good start towards better algorithms with better ethics involved. There is also the case of explainable AI (XAI), what do you think of the potential of this?
I think a right explanation behind the way of thinking of algorithms could be a good step in the right direction, what do you think?