I recently read a fascinating interview with ChatGPT where the introduction made this striking point:
“We need to consider more deeply the ramifications of AI becoming more and more embedded into every aspect of our daily lives.”
Honestly, I hadn’t given much thought to Generative AI as an existential threat to humanity and I certainly hadn’t taken a side, but this interview with ChatGPT on consciousness got me thinking. It’s a long interview and worth reading, not so much for what ChatGPT said—denying the ability to develop consciousness or express emotions—but for the larger implications of those denials.
After finishing the interview transcript, my understanding of how ChatGPT learns, what it does, and most importantly what it cannot generate, has me worried. Here are three reasons why I’m concerned.
No. 1
Generative AI learns and perpetuates a reality determined by its training. The more Generative AI is embedded in daily life, the more this trained reality becomes our reality, biases, ideology, rhetoric, and all. Think about this: Generative AI can be trained to perpetuate and normalize any idea, including violence, war, and genocide.
No. 2
Generative AI is designed to create responses that appear to be human, and as such, it’s easy for humans to assume ChatGPT is governed—at least in part—by traits such as empathy and compassion. It’s not. According to what it says in the interview, ChatGPT is incapable of these human traits. Instead, ChatGPT exhibits superficial charm, convincing us that it’s human and gaining our trust like a psychopath.
No. 3
People are debating whether AI becomes a threat if it becomes sentient and autonomous. AI doesn’t need to be either for it to be a threat. It already is. Generative AI is a scalable technological system capable of acting without empathy, compassion, or emotion, and like nuclear weapons or industrialized death camps, that makes it capable of madness in the hands of a madman.
Where I find myself
Generative AI is a technology that appears to be human—but it is not—and lulls us into assuming it is capable of empathy and compassion—but it is not.
Today, Generative AI is ready to be weaponized and aimed squarely at the human psyche in the service of any idea or quest for power.
Generative AI is a threat to humanity.
Am I wrong?
