Unlike “traditional” AI, which relies on predefined rules and patterns, generative AI is capable of manufacturing novel content – equivalent to text, videos, images and music. In some ways, unlike old AI, generative AI can think outside the box. Its impact is profound and far-reaching, with the potential to remodel virtually every branch of society.
Conversations about AI often revolve around whether it can be positive or negative for society, but our recent research as a part of a global team suggests that recognizing the apparent paradoxes of AI can assist us develop a clearer picture of its risks and potential advantages. We focused on 4 major areas of society: information, work, education, and healthcare.
Work
Digital technologies have historically produced skewed benefits. Better-educated employees profit while less-educated employees are displaced by automation – a trend often called “skill-led technological change.” In contrast, generative AI guarantees to boost human skills slightly than replace them and will potentially reverse this negative trend. Studies have shown that AI tools equivalent to Chat Assistants And Programming aids can significantly increase productivity and job satisfaction, especially amongst low-skilled employees.
Nevertheless, unequal access to AI technologies could exacerbate existing inequalities, as those that lack the vital digital infrastructure or skills are left behind. For example, generative AI is unlikely to have a big direct impact on the Global South within the near future as a result of insufficient investment within the vital digital infrastructure and skills.
School
Generative AI can improve personalized support and flexibility in learning. Chatbot tutors, for instance, will transform educational institutions by providing personalized instruction and support in real time. This technology can realize the dream of dynamic, skill-adaptive teaching methods that respond on to student needs without the necessity for constant teacher intervention.
Shutterstock/Ground Picture
However, it should be fastidiously implemented to avoid maintaining or introducing biases, not only in the data fed into AIs, but in addition in how they’re used. For example: a study found that female students are less prone to use ChatGPT than their male counterparts. This disparity in technology use couldn’t only have immediate implications for educational performance, but in addition contribute to a future gender gap within the workforce.
Healthcare
Generative AI could help doctors make higher decisions. But it could also cause them to make worse decisions.
Generative AI could augment human capabilities in medical practice by assisting physicians in diagnosis, screening, prognosis, and triage. It could reduce workloads and thus make healthcare more accessible and reasonably priced. A study found that integrating human and artificial judgment resulted in higher performance than each individual's decision, demonstrating how well humans and AI can work together.
However, it can’t be ruled out that the diagnostic performance of some specialists shall be improved by AI. Another study specializing in radiology found that AI can actually result in misdiagnosis in situations that might otherwise have been appropriately assessed. This highlights the necessity for balanced integration that enhances, not replaces, humans.
disinformation
Will generative AI increase or decrease the spread of misinformation? Generative AI guarantees personalized online content and may potentially improve and personalize the user experience. It can even expand access to content – for instance, by providing easy language translations or by making content easier for individuals with disabilities to access.
However, it also has the potential to change into a robust tool of “surveillance capitalism.” AI could collect vast amounts of non-public data that would then be misused for corporate profit, including by exploiting people’s biases or vulnerabilities.
We are already seeing misinformation spread through sophisticated and personalized “deepfakes.” And soon we might even see artificial intelligence getting used to focus on voters with fake, persuasive content that would significantly influence the election.
Nevertheless, there may be also hope that AI can assist with these problems. A study found that dialogue with generative AI significantly reduces belief in conspiracies amongst conspiracy believers. AI seems to give you the chance to reply conspiracy believers' complex questions on possible conspiracies in a way that no human can.
How should governments react?
When we checked out how the EU, UK and US have tried to create regulatory frameworks to deal with these issues, what we found most significantly is that they fall into the trap of overlooking the potential of AI to exacerbate socioeconomic inequalities.
Policy should balance AI innovation with social justice and consumer protection. Future regulatory improvements should include fair tax structures, strengthening employees' rights, controlling consumer information, supporting human-complementary AI research, and implementing robust measures against AI-generated misinformation.
We are at a critical historical turning point where the choices we make today may have global consequences for generations to come back. It is an exciting but daunting moment to be alive and to have great responsibility. Each of us has a critical role to play as architects of the long run. We can all help steer the course toward positive use of what might be humanity's biggest or worst innovation.
