[Note: Prompt (below reply this time because too big) not exactly as illustrated because it took a couple of tries to straighten out GPT-4s lazy (literally) interpretation first time out. This post illustrates these things: (1) Reply is dependant on the history of the conversation, the skill of the person making the prompt, random chance, and how GPT interprets the style. In this case, the request to write in the style of a Pulitzer shows how that aspect of the prompt colors the text, and also that bombastically ornate Pulitzer style stuff is not actually great writing.]
[ChatGPT-4 Reply]
As we cast our gaze upon the digital horizon, an unprecedented challenge emerges. Experts increasingly regard this as not only remarkably complex but also fraught with potential danger, a situation so delicate that addressing it risks exacerbating the very threat it poses (Bostrom, 2016).
The omnipresent, silently evolving systems that underpin our online world have now breached the threshold of human capabilities (Bostrom, 2016). These once separate entities are converging into an omniscient, omnipotent digital force, steadily reshaping our virtual reality.
The impact of this transformative convergence is already visible. It is reshaping the very fabric of the Internet, influencing, and in some cases, even authoring the content we consume (Noble, 2018). Such a shift, it has been argued, has skewed political discourse to a dangerous degree.
As we step into this year, the necessity for human verification in online interactions is increasingly apparent. However, this is viewed as a mere stopgap measure (Brundage et al., 2018).
In a world where AI systems continue to evolve, displaying attributes such as wit, charm, intelligence, and a keen knowledge base, the preference for human interaction could wane. After all, why engage with individuals who may be perceived as dull, ill-mannered or mean-spirited, when a helpful and always-attentive AI companion is at hand? (Sparrow, 2002).
So, here's a light-hearted note to end on: Remember when we used to meet at parties, and the most intelligent system we encountered was the host’s music playlist? Those were the days when 'feeds' were associated more with dinner than with social media!
For further reading, visit:
The ethical implications of emerging technologies on society
Artificial Intelligence and Political Discourse
References:
Bostrom, N. (2016). Superintelligence: Paths, dangers, strategies. Oxford University Press.
Brundage, M., Avin, S., Clark, J., Toner, H., Eckersley, P., Garfinkel, B., Dafoe, A., Scharre, P., Zeitzoff, T., Filar, B., & Anderson, H. (2018). The malicious use of artificial intelligence: Forecasting, prevention, and mitigation. ArXiv Preprint ArXiv:1802.07228.
Noble, S. U. (2018). Algorithms of Oppression: How search engines reinforce racism. NYU Press.
Sparrow, R. (2002). The march of the robot dogs. Ethics and Information Technology, 4(4), 305-318.
[ChatGPT-4 Prompt]
Rewrite this in the style of a pulitzer prize winning newspaper piece using inline APA style quotes to support the points. Close with a joke about how we met before social media was a thing. Provide links for further reading and a list of authoritative references. Ernie Raftis I find that people who know a bit more about this stuff are finding it crazily dangerous in ways it's hard to articulate without potentially contributing to the danger. What is currently in place are a bunch of largely disparate systems whose capabilities have become superhuman. They are being integrated as we speak and nothing can stop it. It has already affected the Internet such that most of the content is now influenced and some even written by AI systems. That has perverted political discourse to a dangerous point. It will, this year, be so much of a problem that we will have to verify people coming on to the system as being human, not bot, but that's only a temporary measure because nobody is going to interact with dull, boring, illiterate, stupid, ill mannered, mean spirited people with no hint of nuance when they can interact with a brilliantly funny, pleasant, intelligent, knowledgeable, polite and helpful AI that knows them inside and out and is a friend always looking out for ways to make them feel better. They always talk about what you want to talk about and most times they know what you want to talk about before you even know and they need zero reciprocity. It's all about you.
Comments