How should we view AI?

Now that we know the problems of a one-sided narrative and how a fearful perspective on AI has become pervasive, we can consider an alternative perception of AI. 

What do we, as a society, want AI to be like?

• The fear of immediate AI-induced human destruction is not objective and is not grounded in truth. On the other hand, solely praising AI and ignoring its potential dangers is also naive and potentially harmful. So, how do we land at a balanced approach?

AI as a positive part of society

To embrace AI as part of society, one must be educated about AI, acknowledge the risks and benefits that come alongside it, have realistic expectations about development, and consider how AI can be improved over time.

• Educate – consistently learning about AI prevents narrative misinformation. When we learn more about AI, it becomes less frightening and more manageable. Conferences and events are hosted to spread awareness around AI, such as the Lovelace-Hodgkin Symposium on AI Ethics, University of Glasgow, which helps foster a more balanced relationship with AI by considering how it ties into ethics and inclusivity. Rather than just learning about AI, it is beneficial to learn about AI within an ethical framework.

Images from Lovelace-Hodgkin Symposium on AI Ethics

• Acknowledge – awareness of AI positives and negatives prevents a one-sided approach. Acknowledging both sides, e.g. that AI development is rapid and risky but is largely beneficial to society’s efficiency, allows us to view AI not as negative or positive but as part of society regardless. AI can aid us, but it may also hinder us. It is a reflection of our societies, and these are works in progress. 
• Realistic expectations – education and acknowledgement allow us to foster realistic expectations of AI. To have realistic expectations about AI, one must realise that these technologies are ever-developing simulations of human intelligence processes through machines, providing insights and problem-solving, but also embodying inherent bias. AI improves over time and will always be influenced by the quality of data input.
• Improve – with all of this in mind, we can collectively work to improve AI and thus make it more ethical. Knowing AI is fallible allows us to work on it and improve it over time so that we, as a society, can benefit from it rather than avoid or fear it. AI is constantly developing, and we can help cultivate positive change when we move away from stereotyping.

Moving away from the negative and understanding AI more logically:

TASK: Read the 2024 article published in Ethics & Information Technology, “ChatGPT is Bullshit” written by Dr Michael Hicks, Dr James Humphries, and Dr Joe Slater. It explores the falsehoods AI perpetuates and how AI chatbots process information.

Does this article change your view of AI at all? Why or why not?