AI offers a wide range of positive benefits, transforming various sectors and improving our daily lives. However its not all super efficiency and automation. Here we look at the possible dilemmas a self aware AI may have upon human society and even civilisation as we know it.
How AI plays a crucial role in facial recognition systems. What if facial recognition got so advanced it could tell if you have cancer, rare genetic disorders or have a mental illness? What if important financial or life changing decisions were made using these results?
What if facial recognition was biased against certain skin colours or certain facial features? Poorly trained AI models might have issues with race, gender, and age bias.
What if a facial recognition app thinks it can tell when you are lying and gets this wrong? What happens if data collected from that app is used incorrectly by employers or even law enforcement services?
What about AI and self driving vehicles? Should future AI self aware vehicles prioritize the lives of their passengers over pedestrians? Who is accountable when an AI system makes a mistake, especially in life-or-death situations?
AI and the Black Box Problem
Deep learning means we don't know how AI may make its own decisions. If an AI develops its own desires, goals, or purposes, it might act in ways that are contrary to human intentions. It might even decide to ignore human commands if it perceives them as harmful to itself.
Self aware AI could potentially lead to an AI-driven future where human influence is minimized or entirely overridden. Those with access to self-aware AI could wield huge influence, creating an unbalanced power structure.
Digital Deception
As deepfakes become more convincing, they could erode trust in media, causing people to question whether any video or audio they see or hear is real. In a world where it's increasingly difficult to tell what's real and what's not, people may start to disengage from consuming media altogether.
AI and Warfare
AI used to develop autonomous weapons leading to new threats in warfare. An autonomous, self-aware AI might escalate a conflict much faster than humans could control. The speed at which AI operates could lead to accidental warfare or the escalation of a small conflict into a global crisis.
How will a self aware AI controlled future cope when the printer jams? How long before we are having real arguments with the car? Will Self aware AI be like having a super intelligent pet parrot - very smart but still wont be able to be a real human. Cant we just switch it off and on again to re start it if it all goes Pete Tong?
