AI has brought concerns that have generated uncertainty and preoccupation in society. In this post, some questions will be answered to reduce this uncertainty and find ways to mitigate future problems.
Will there still be jobs?
Yes. As Acemoglu et al. (2020) highlighted in their study, AI does not significantly impact employment or wages. However, AI-exposed companies have shown a reduction in hiring non-AI positions and a change in the skill set demand aimed at AI-related skillsAI will displace some jobs and create new and more specialized ones (Acemoglu & Restrepo, 2017).
Will AI generate more inequality?
Yes. Although AI has been demonstrated to improve people’s life quality, I found three factors that, if not addressed, will increase economic inequality:
- a) Disparity between small and big companies: Big companies are more likely to have the required resources to implement better and more competitive AI to take advantage, creating a more significant gap between them and much smaller companies that may end up out of business due to the lack of competitiveness.
- b) Specialized education needed: As mentioned in the first numeral, AI-exposed companies require more specialized roles, which require better and more expensive education, which is inaccessible to people with limited resources. This will push them to mid-skill level jobs with lower wages.
- No increase in wages: AI has brought multiple benefits to companies adopting these technologies, including increased productivity and better resource allocation. However, these augments in productivity have not translated into an increase in wages for employees, meaning that most of the benefits from AI are going to the business and not its workers, thus increasing inequality.
Will a few large companies control everything?
We cannot know if this will happen since we cannot predict the future, but a few large companies are likely to take control of the entire AI scene. Not all companies have the resources and talent to keep up with the evolution of AI, and for new starting companies, it is even more challenging.
Limitations will lead smaller companies to be sold to larger companies or liquidate themselves (winner takes
all).
Will countries engage in race-to-the-bottom policy-making and forfeit our privacy and security to give their domestic companies a competitive advantage?
This is more likely to happen in poorly legislated countries and authoritarian regimes, usually in developing and underdeveloped countries. The use biometric and behavioral data to
identify citizens considered threats to government interests, violating people’s privacy to influence their decisions are some of the risks of poor regulations (Hyman, 2019). We must demand our leaders to create regulations and ensure they comply.
Will the world end?
There are several risks and challenges associated with AI use. But they can be prevented with robust regulations aimed to protect people’s privacy and security.
Other risk may arise from poorly defined objectives, that may lead AI to commit unethical or even illegal actions in order to achieve these objectives. With techniques such as Inverse reinforcement
learning, we allow machines to learn from humans and stick to their values and ways (Gulchenko, 2024).