False Idols category archive

Fomenting the phony. Well, maybe just a bit. From El Reg:

Originality.AI, an online AI detection platform, examined 8,795 long-form LinkedIn posts – defined by the site as 100 words or longer – and found that, based on its own assessment tools, 54 percent of such posts “are estimated to be AI-generated.”

This trend suggests that many self-proclaimed thought leaders on LinkedIn may be presenting AI-generated content as their own profound insights.

Much more at the link.

Aside:

Speaking of the land of the me-me-me . . . .

Via C&L.

At Psychology Today Blogs, The Open Minds Foundation takes a lot at the potential effects of AI-generated dis- and misinformation on the internet. They conclude that internet users need to exercise more critical thinking skills, even as they seem to be exercising less (or is it fewer?).

Here’s a tiny bit from their article; I urge you to read the rest.

Psychologists at the University of Cambridge recently developed the first, validated “misinformation susceptibility test” (MIST), which highlights the degree to which an individual is susceptible to fake news. Younger Americans (under 45) performed worse than older Americans (over 45) on the misinformation test, scoring 12 out of 20 correctly, compared to 15 out of 20 for older adults. This was in part correlated to the amount of time spent online consuming content, indicating the relevance of how you spend your recreational time.

The Europol report continues with a stark warning: “On a daily basis, people trust their own perception to guide them and tell them what is real and what is not… Auditory and visual recordings of an event are often treated as a truthful account of an event. But what if these media can be generated artificially, adapted to show events that never took place, to misrepresent events, or to distort the truth?”

The Learning and Implicit Processes Lab at Ghent University takes a look at the current state of ChatGPT (and Large Language Models in general) and concludes (emphasis added):

Returning to the question of ChatGPT’s intelligence, it is important to note that it was developed with a specific purpose: to interact with humans through a computer interface and produce coherent answers to whatever prompt it gets. With that goal in mind, its performance is remarkable.

It was not designed to be generally intelligent (i.e., capable of flexibly adapting to novel situations or problems), and it isn’t. Still, it gives us the illusion of intelligence because it mimics intelligent human language.

Follow the link for their reasoning.

ChatGTP is getting a lot of play lately. It even made an appearance in my local rag’s comics page.

Before mucking about with it, I suggest that you listen to Harry Shearer’s interview with Gary Marcus on the January 1, 2023, edition of Le Show; the relevant portion starts at about the 20 minute mark. Also read this from Bruce Schneier’s website.