In today's ZOOM session, I mentioned the idea of a central executive for an AI system. As a starting point, imagine something like Siri, Alexa, Cortana -- but with much more smarts. AI systems, both new and old, can provide a large part of the smarts. Like humans, they wouldn't be infallible. But they could support a central executive that would be held responsible in case of errors or problems or disasters.
The critical issue is RESPONSIBILITY. The central executive would be comparable to the CEO of a corporation. EVALUATION is essential. The central executive, like the CEO of a corporation, would know how to get any info that may be needed and who could evaluate it against whatever business, legal, factual, or ethical criteria are critical.
LLMs are very good for finding information in ordinary language. But they are not good at evaluating that information. More traditional AI reasoning systems are more accurate and more reliable. The central executive must have both kinds of abilities -- supported by appropriate AI assistants.
I attached Section7 of an article I recently finished. It explains some background from psychology, neuroscience, and computer systems. See especially Figures 18, 19, and 20.
Figure 18 represents human reasoning. Figure 19 shows how an AI central executive could play the role of a human. and Figurer 20 shows a similar "OODA" loop that has been used to analyze and solve “wicked” engineering problems, which involve “complex interdependences" between the systems and incomplete, inconsistent, information about the problems.
And see some excerpts below from an earlier note I sent to Ontolog Forum.
John
____________________________________
Sent: 5/5/24 9:52 PM
The psychologists Baddeley & Hitch wrote their initial article in 1974. They wrote in response to George Miller's hypothesis about the "Magic number 7, plus or minus 3". They realized that there is much more to short-term memory than just words and phonemes. They called Miller's storage "the phonological loop" and they added a "visuo-spatial scratchpad" for short-term memory of imagery and feelings. And they continued to revise and extend their research for another 20 or 30 years.
Neuroscientists, who are specialists in different aspects, have been working on related issues. The consensus is not a single hypothesis, but a branch of research on issues related to conscious control of action by a central executive in the frontal lobes vs. subconscious control by the brainstem and the cerebellum.
For example, when you're walking down the street and talking on your cell phone, several different systems are controlling your actions: (1) the central executive is in charge of what you're doing on the phone in talking and pushing buttons; (2) the cerebellum is guiding your steps in walking and maintaining your balance; (3) the brain stem is maintaining your breathing, heart beat, and other bodily functions; and (4) the nerves running done the spine and branching to all parts of your body are controlling every movement and monitoring any abnormalities, such as a burn, a scratch, or a more serious injury.
In Freud's terms. the central executive is the ego, and the lower-level systems are the id. Those ideas are much older, but they illustrate the kinds of issues involved. The more recent research relates the observational data to actual neural functions in specific regions of the brain. Since aspects of those functions can be traced back to the earliest bacteria, worms, and fish, there must be something fundamental about them. AI systems that do not support related functions do so at their peril.
In my notes and the articles I cite, there are many references to ongoing research. For more background, don't use those GPT-based things that summarize surface-level trivia. You can start with Wikipedia, which cites the original research. Then continue with more detailed studies in neuroscience.