Why Should Be Surprised?

I read an article in Substack about "Latent Knowledge." From what the author wrote, I couldn't tell whether he wanted to be cited for the term or someone else came up with it. No other articles were referenced in this author's work, so I shouldn't be surprised.

If this author knew anything about AI (really about its history) and more than that about its terminology, he would have known about something called "Knowledge Management." In the 1990s, this was an effort to establish a means to leverage the commodity we call knowledge so that organizations can collect it, organize it, and use it for their own benefit.

Although today, the field of knowledge study and management has more or less receded into the background, it is my belief that it is still extremely relevant to everything we do. After all don't we use knowledge and information on a regular basis. Although in today's world this might seem somewhat questionable.

With the effort to create these massive databases of statistics that somehow represent knowledge (they do not), the past is forgotten mainly to them, especially all of the research done to codify knowledge so that a computer can process it. Humans could understand something about how that was accomplished.

The current crop of AI researchers seems to have no concern about such approaches. It appears that their approach to solving the problems of intelligence is to scrape larger and larger datasets with the hope that somehow AGI will magically arise one day from these statistical databases.

AGI will never magically emerge from these databases. These artifacts have no basis in human intelligence. At best, they simply represent an approach to getting answers from a relatively stupid and unreliable oracle (the LLMs).

Back to the article that I started with. Had the author of that article considered anything that was done in the past, they might have realized that there is a concept defined under the auspices of the study of knowledge called TACIT knowledge. Here is a definition of the term. (I will cite my sources.)

The following definition is from Wikipedia (https://en.wikipedia.org/wiki/Tacit_knowledge)

"Tacit knowledge or implicit knowledge—as opposed to formalized, codified or explicit knowledge—is knowledge that is difficult to express or extract; therefore it is more difficult to transfer to others by means of writing it down or verbalizing it. This can include motor skills, personal wisdom, experience, insight, and intuition."

Doesn't this sound like what our author friend calls "latent knowledge?"

If that definition doesn't convince you, the next one will provide further reinforcement that we don't need to invent a new phrase for something already defined.

"Tacit knowledge is the knowledge you've gained through living experience, both in your personal life and professional development. It is often subjective, informal, and difficult to share or express because it is affected by our individual beliefs and values, from: https://document360.com/blog/tacit-knowledge/. "

I've said enough to get my point across.

A Brief Note

I regularly read ARTICLES about AI as they appear in newsletters, published papers etc., etc. A recent article entitled "Microsoft's Mustafa Sulyman says he loves Sam Altman, believes he's sincere about AI safety." I will be writing an article about this story sometime soon. I'd like to give y'all a heads-up about the article. Mustafa is PRO Sam Altman.

Like what you've read? Then subscribe. That will tell me for sure that you like what you read.

Thank you.

Dr. Randy M. Kaplan

PLEASE NOTE: This article was not written by an AI. I wouldn't dare.