Generative Artificial Intelligence or Generative AI is an innovation with a terrible name.
First, generative means something that originates, produces, or even reproduces.
Next, artificial Intelligence is defined as intelligence exhibited by machines (machine intelligence).
So, if the two are put together the result is a machine that originates intelligence—in other words, a thinking machine. Hmmm!
Generative artificial intelligence is typically defined as a technology capable of generating text, images, videos, or other data, usually in response to prompts.
At best the name generative AI is confusing, and to many it is scary.
Everett Rogers in his classic book Diffusion of Innovations (2003) states that “the name given to an innovation often affects its perceived compatibility, and therefore its rate of adoption (p. 250).”
Certainly, Generative AI is an innovation—it is new and is probably the innovation of this decade.
Everett Rogers defined Innovation as “an idea, practice or object that is perceived as new by an individual or other unit of adoption (p. 12)”.
What does all this mean to distance educators?
Maybe first we should learn more about Generative AI. So, just for the fun of it, the written sections above were entered into Gemini AI, a generative AI tool, and a competitor to ChatGPT.
Here is what Gemini produced:
You’re right, “Generative AI” while descriptive, and can be misleading.
Here’s a breakdown of why it might not be the best term and some suggestions for alternatives:
The Problems with “Generative AI”
Misrepresents AI’s role: Generative AI doesn’t create intelligence; it uses existing data to create new content that resembles what it has learned. AI itself doesn’t possess true sentience or independent thought.
Focuses on creation, not the process: It emphasizes the outcome (generation) but doesn’t explain the underlying method (learning patterns from data).
Alternative Terms
Machine Learning for Content Creation: This clarifies the role of AI (machine learning) and the goal (content creation).
Creative AI: This emphasizes the creative aspect of the technology.
Data-Driven Content Generation: This highlights the reliance on data for generation.
The best term will depend on the context. “Machine Learning for Content Creation” might be more suitable for a technical audience, while “Creative AI” could be better for a general audience.
Additional Points
The reference to Everett Rogers’ “Diffusion of Innovations” is spot on. Clear and concise naming can significantly impact how readily people accept a new technology.
While Generative AI might be confusing at worst, it’s not necessarily scary. It is best to focus on the applications and benefits of the technology to help alleviate fears.
This revision process took about 5 seconds, and by any measure the results were impressive—the Generative AI analysis seemed insightful and accurate.
Rogers recommends that potential innovation users should learn about the innovation by studying its attributes—there are five—relative advantage, compatibility, complexity, trialability, and observability.
First is the idea of relative advantage—defined as the degree to which an innovation is perceived as being better than the idea it replaces. Relative advantage is often expressed as the innovation’s economic profitability or its ability to convey social positions.
The next characteristic of an innovation is its compatibility, which is the degree to which the innovation is consistent with existing values, past experiences, or user needs. Innovations can be either compatible or incompatible.
Next is complexity, explained as whether the innovation is perceived as difficult or easy to use.
A complex innovation, as perceived by potential adopters, can significantly hinder adoption.
Trialability is the ability to experiment with an innovation before adoption. If an innovation can be easily tried it will be likely to have a rapid rate of adoption or rejection.
Observability is the visibility of applying the innovation. Observability means seeing the results of using the innovation. High observability promotes faster decisions about adoption.
In other words, does Generative AI allow us to do things better? Next, are the results of use compatible with what the user needs or wants? Third, is Generative AI easy or difficult to use, and can we quickly try it out? Finally, can we see the results when Generative AI is used?
Computer scientists argue that Generative AI is better, compatible, easier, not complex, and results are clear—maybe this is true, but the name is still horrible.
Generative artificial intelligence sounds intimidating and threatening. Perhaps ‘creative artificial intelligence’ would be more inviting—perhaps not!
Distance educators should understand, study, and evaluate Generative Artificial intelligence and write about it—perhaps someone from the U.S. Distance Learning Association could “coin” a new name.
(Distance Learning would love to publish manuscripts on this and other related topics.)
And finally, Galen said “The chief merit of language is clearness, and we know that nothing detracts so much from this as do unfamiliar terms.”
NOTE: Rewrites of the second half of this column using personal intelligence (PI) took 5 tries and 4 hours—a relative advantage?
About the author
