The absence of digital competencies and access to online services has real costs for those excluded. In this article, we will briefly discuss the impact of unreliable internet connections, the hidden costs of outdated equipment and the replacement of entry-level jobs through AI-based systems.
Slow and unreliable internet connections might mean that individuals are not able to participate in video calls. It also means that the individual’s productivity is much lower as time is wasted to load pages or download files. Furthermore, creating content such as videos for platforms becomes a challenge.
There are also hidden costs when it comes to the equipment.
Researchers led by Tobias Berg from the Frankfurt School of Finance & Management found that, for example, car drivers applying with a “Hotmail” e-mail address for car insurance are charged higher fees, as the insurance company finds that some e-mail domain names are associated with more car accidents than others. Across the finance industry, credit scoring models even use the phone battery life, the time of the day, browser settings, spelling errors, or the font installed on the computer.
It is not hard to see how the mechanisms governing the digital space negatively impact those who have lower digital skills and even lead to negative feedback loops.
A person’s limited digital skills can influence their income. In addition, tracking models may notice that the process of entering personal details takes more time and those clients should be charged higher fees for loans or insurance policies. These elevated fees result in increased living costs, which could potentially necessitate relocation to another postal code area. Postal codes are also frequently used for customer classification, which might, once again, lead to higher fees for them to pay.
While profitability remains crucial for any business, it is imperative to acknowledge the presence of these loops and biases inherent in economic systems.
Another domain is the emergence of Generative AI models such as ChatGPT, Bard, Midjourney or DALL-E. Many existing entry-level jobs are thought to be taken over by AI-based systems. These large language models are already used for customer service, phone reception, translation, basic content creation or data entry among many use cases. In addition, workflows are becoming ever more digital and require employees to be familiar with these tools.
So, what steps can we take? It appears that governments ought to assume more substantial roles in addressing algorithmic bias, while regulators should explore collaboration with companies and the adaptation of current frameworks to encourage the development of ethical algorithms. Moreover, in the context of AI’s impact on the workforce, particularly for less-skilled employees, governments could consider nudging AI companies to develop products aimed at enhancing rather than replacing workers without losing competitiveness. For instance, AI that offers research assistance to TV writers without generating scripts, as the latter might likely be of lower quality.
Wolfgang Spiess-Knafl, Giulia Parola
 Berg, T., Burg, V., Gombović, A., & Puriand, M. (2020). On the rise of FinTechs: Credit scoring using digital footprints, The Review of Financial Studies, 33(7), 2845–2897. https://doi.org/10.1093/rfs/hhz099
 Goldberg, E. (2023, May 23). A.i.’s threat to jobs prompts question of who protects workers. The New York Times. Retrieved from https://www.nytimes.com/2023/05/23/business/jobs-protections-artificial-intelligence.html.