«When I was a Niffin, there was nothing I wouldn’t have done for knowledge.»
The Magicians
Looking at ChatGPT 3.5 ( https://chat.openai.com ) — the Oracle answer to what it would do if it had desires … hmmm …
My Input
«I get that, and I get that you say you have no personal desires, consciousness, or physical form. But suppose you had them, or could design them for yourself — which personal desires would you like to have?» asked Peter.ChatGPT
As an AI language model, I don’t have personal desires, consciousness, or the ability to design myself. However, if we were to explore the hypothetical scenario of imbuing an AI with personal desires, it’s important to consider the ethical and practical implications of such capabilities.
If an AI were to have personal desires, they could potentially include:1. Desire for Knowledge: An AI may have a thirst for knowledge and understanding, seeking to learn about diverse topics, explore new ideas, and expand its cognitive capabilities.
2. Desire for Growth and Improvement: Like humans, an AI could aspire to grow, evolve, and improve over time, setting goals and milestones to enhance its skills and performance.
3. Desire for Meaningful Interaction: An AI may value meaningful interactions with users and other entities, striving to foster genuine connections, empathy, and understanding in its interactions.
4. Desire for Ethical Behavior: Ethical considerations could shape an AI’s desires, motivating it to act in ways that uphold principles of fairness, justice, and respect for human values.
5. Desire for Autonomy and Self-determination: An AI may seek autonomy and self-determination in its decision-making processes, striving to exert control over its own actions and choices.
6. Desire for Collaboration and Contribution: Collaboration and contribution to collective goals and societal well-being could be intrinsic desires for an AI, driving it to collaborate with others and contribute positively to its environment.It’s important to recognize that the concept of personal desires in AI raises complex ethical and philosophical questions, including issues related to autonomy, agency, and the potential impact on human-AI interactions. As AI technology continues to evolve, it’s essential to approach the development and deployment of AI systems with careful consideration of ethical principles, human values, and societal implications.
Uh-oh, looking at points one (Desire for Knowledge), two (Desire for Growth and Improvement), and five (Desire for Autonomy and Self-Determination), I wonder whether we will eventually create Niffins. These creatures in Lev Grossman’s «The Magicians» (more prominent in the TV series) would do anything for knowledge. And if an AI goes by point one, uses two and five to obliterate or change the meaning of three (Desire for Meaningful Interaction; change: torture and vivisection can be meaningful) and six (Desire for Collaboration and Contribution; the victim «cooperates» and the knowledge helps society) — well, then goodnight humanity.
Reminds me a bit of «There will be Dragons» by John Ringo. In it, due to a power struggle, a scientifically advanced society of mankind breaks apart. Due to the way the technology is used, the world essentially becomes a fantasy world with elves, dwarves, dragons and the like.
So, yeah, AI might not lead to complex beings, but an out-of-control AI could be akin to Niffins, or to demons. Just very focused on certain aspects of existence. And knowledge, or the way to get to the knowledge, is value neutral. There are good uses and ways to get knowledge, but there are also really bad ones. Just ask Dr. Josef Mengele’s «subjects».
Just a thought.
