The contents of Volumes 1 and 2 are a combination of the human author's conceptual work (coming up with the core ideas, the overall thrust of the content, the structure, and the intellectual landscape) and the text generation by one of more of these LLM models:
- Claude Pro AI
- Gemini
- Perplexity AI
- GPT-4
- xAI Grok
- Copilot
These LLMs were tasked in the creation of the contents of this volume to put the flesh on the bone-the basic idea content and structure of each piece in this miscellany. They did so as I directed them to perform in the generation of text, tabular data, and graphics.
All the LLMs performed as LLMs perform-without any consciousness of what they were about when engaged in one or more exchanges with me.
Being coded constructs trained on vast curated troves of data, these LLMs were merely dumb but obedient assistants in helping me to put together all of the artifacts in this collection that conveys to you, the human reader, what I have been thinking about in connection with the basic cause and effect that comes about because of the activities of two groups of humans: 1) those who create the LLMs (coders, linguists, computer scientists, algorithm designers, ethicists, and: 2) all the rest) and those who use them, like me, an 82 year old Doctor of Philosophy who has spent quite a bit of time in the past two years thinking about LLM models, what they mean, and how they should and should not be used-that is, what are their limits and how can humans make use of them without giving the coders too much control on the output.
[LLMs will never write a paragraph-length compound, complex sentence. That's what humans do! ]
On balance, I find there are many problems facing academia when it comes to trying to devise ways to control the use of LLMs in all areas and phases of academic research and the creation of scholarship that is submitted to peer review for publication in journals of all kinds on a global scale.
One major problem are the many different business models that now define the intersection of LLMs with academic publishing of all kinds, again on a global scale. Citation counts are the coveted metric, the analog to click bait on so many sites online (YouTube and the like).
Another is the inability of academia to come to grips with the reality that LLMs are not going away, that they serve a positive purpose in the pursuit of human knowledge, and that scholars need clear guidelines about what they can and cannot do when using one or more LLMs in their scholarly research and the process of creating their papers and submitting them to reputable journals for publication after being peer-reviewed by scholars (who may or may not also use LLMs in the peer review process).
Finally, the inability of academia to find its footing with respect to the continuous advancements in the LLM world is harming bona fide use of LLMs as powerful assistants to the human scholars across all knowledge domains who are curtailed in their responsible and ethical use (
nous and
phronesis via
praxis) of LLMs in the pursuit of
eudaimonia. My hope is that the changing demographics in academia will help to bring about ethical guidelines covering all facets of using LLMs in research and scholarly publication.
The result might well be that the publishing world will learn to accommodate and fold into its evolving business models what is best in the human/LLM dynamic.
It is my belief that, sooner than many think, the widespread and accepted use of LLM models in both academia and scholarly publishing will become a commonplace reality.
As long as that reality is based on the Aristotelian triad of
nous, phronesis, and
praxis, there is much to be gained (
eudaimonia) by the ethical use of LLM models across academia and its output in the way of scholarly publications.
Sign-In
Cart