The chances of finding out what’s actually going on are so absurdly remote that the only thing to do is to say, “Hang the sense of it,” and keep yourself busy. I’d much rather be happy than right any day.” ~ Douglas Adams
This is a time when, despite Douglas Adams wise take on ‘Life the Universe and Everything”, we do have cause to ask the “Big Questions”. But are answers any more available than previously? When we ask what education and learning is for in the age of ubiquitous AI, as we now must do, the big questions about what is human life “for” arise; what is our destiny, and what do we need to know to get there?
If it’s not to prepare to be a part in 20th century economies, and instead is preparation for some yet undefined knowledge work in the 21st century, what knowledge are we talking about that students need to learn? And is it knowledge that isn’t going to become obsolete through replacement by AI entities, neural black boxes, and “all knowing” algorithms?
We might take a look at some aspects of how our current “knowledge tree” arises; perhaps there is some guidance there for what learners need to learn going forward. Where does knowledge come from?
-
- Personal experience
- Accumlated experience in a group, a culture, a people
-
- Oral transmission
- Written transmission
- YouTubes (streaming video and social media formats)
-
- Specialized knowledge defined in disciplines (silos) such as scholars pursue through years of learning and higher education, Doctorates and post doctorate research, usually deriving consensus with others in the same discipline.
- Currently we are not really able to take knowledge from one discipline and use that to “know” a different silo: Art and Science, History and Math, apples and oranges.
- AI
- Training on existent human knowledge sources including the totality of “the internet”, involves synthesizing some new form of a “knowledge tree” that we know is there from the AI outputs, but don’t truly understand how such vast amounts of knowledge are restructured into much much smaller sized “containers” or LLMs which remain available to query.
- Advanced AI ability to think, reason, and distill vast amounts of knowledge across disciplines and silos, taking what was formerly known and deriving and “creating” new knowledge that supercedes the former “best thinking” of human knowledge trees.
Scientific knowledge is generally thought to be a unitary undertaking in that one type of science fits into the structure of the same reality as another type of science.
Yet this is seemingly not universally so, as in the challenges of Physics where we have one model for a certain scale of reality, and a different model for a different scale… and so far we can’t find the unity between them. One might also consider our attempts to understand the human brain, using in one case the study of evolutionary genetic implementation of DNA, and in another case the electronic signals we can detect in real time coming from brain parts and regions that we attempt to connect to various human behaviors.
IF one posits that psychology and psychiatry also model the brain, or perhaps just behavior that we attribute to brain states, we are using a science that is seemingly different than physics. Perhaps additionally we can ask is biology just chemistry in the human body, or is the chemistry created by the biology.
One might go on with various other examples that may or may not prove useful in the long run, but the essential point is that our best human minds lack some way of distilling science as well as other forms of knowledge into a useful form that transcends our current limitations.
IOW, we would like to know a lot more about life the universe and everything as we attempt to design “machines” that can answer those questions, and that can “teach” students what we need to know going forward.
Another way of looking at all this: take the field we call spirituality or religion or myth or simply ways of knowing that aren’t mentioned in Horatio’s library books. The ineffable world. Much has been orally passed down, and much has been written, but few have been able to find a core to this sort of knowledge. Joseph Campbell made a strong case with “The Hero with a Thousand Faces” that mythology and the sort of explanative narratives we have throughout history can be understood as one core story with identifiable parts that are common to all such.
One is tempted to ask AI what sort of take it might have if it queried Campbell’s amazingly comprehensive scholarship, research, and synthesis. But why would we stop there, as many brilliant minds have toiled in similar fashion creating various religious texts and dogmas, and the range of belief systems beggars belief. There’s literally something for everyone, and if we try to distill or synthesize some essential wisdom about human existence from those sources, it’s way way beyond what any scholarship can provide.
But we do want to know: in human experience of the spirtual or non material aspects of existence (or however one wants to describe extra-empirical knowing) is there some core “voila” moment to be found? Or is it as has been noted, “Elephants all the way down” (the image of what holds up the earth in space) ?
So it follows we’d like to train and input the voluminous human experience and “knowledge” noted above into an AI tool, and see what that AI might be able to derive from it all. What “higher” or more comprehensive level of knowing is there accessible to us? Can we overcome the absurdly low odds that Douglas Adams has given us for the pursuit of what knowledge is for? IF so, perhaps then we can posit what learning is for in the age of AI.