You can ask for information, data, knowledge, or wisdom from a machine. Each of these words has a sense in which the machine’s response will be fitting or not fitting. You can certainly ask a machine to produce two responses, one more meaningful than the other, for a given question, and you will likely recognize the outputs as valid.
If I take your statement literally, it seems to deny the above. Of course, if what you mean is “meaning requires the human meaning maker”, I would agree wholeheartedly. Meaning is like the sound of a drum, the understanding of which requires understanding of both the drum and the drummer (to allude to a famous passage from the Upanishads).
For a fully theoretical treatment of “meaning”, this body of work:
The gist of it is that “meaning” has origins in evolutionary biology. Any organism needs to sort that which it can eat, from that which would prefer to eat it. Signals come in all shapes and sizes. Meaning is signal processing to determine which such signals are actionable and which are not.
IMO, anyone using LLMs for life-guidance is missing the point that for mega-cap technology companies, the customer IS “the meal”. My approach is more like of the hyena - I’ll jump in to steal a bit of the high-tech “kill”, but I do not in any case wish to become the main course! That “means” it’s best to keep one’s own counsel about when to feed off of LLM output when it’s better to try a different watering hole!
Look I’m all for using LLMs as sparing partners. But at the end of the day, it’s our human butts on the line in the arena of life. A person who can’t figure out which move to make, on the street, or in a room, without a phone in their face or getting cues from earbuds, to me just looks like a fat target for anyone more agile than that.
Here’s a use case for LLMs, I wonder if you’d condone it
What we’re discussing presently has veered off topic considerably (based on the title of the thread and original question). It might as well be a completely separate thread.
Sometimes this is innocuous and even desirable, like going off key in a jam session (may lead to something bigger and better). This is analogous to more temperature in LLM speak. Sometimes it can be called ‘hijacking’ (not necessarily in bad faith).
So there’s a conflict here:
we want to discuss certain themes in depth, which requires focus
we want to allow spontaneity and individual expression to flourish
Might a valid use case for an LLM be to interject at some point in a discussion that’s gone off the rails a bit (like this one), in an attempt to get the discussion back on topic?
Or perhaps to inform certain participants that their conduct is not in line with community guidelines?
Such scenarios require only reasonably accurate reasoning ability from a model, and it seems like a legitimate way to resolve the core conflict. As it stands, the conflict goes unresolved.
Circling all the way back to the OP, what’s changed for me since June 25 is I have been reading Whitehead lately, and that puts an entirely different spin on things!
My understanding of Whitehead is that creativity is fundamental to the process of the universe. Whitehead has some technical terms like concrescence and prehension to analyze how entities emerge through processes, but for me (in plain language) the idea of “pulling things together” is a good enough rendition of the Whiteheadian perspective.
OK, so what’s up with the Forum? It’s an opportunity to pull things together. (To “concresce”). From that POV, the “output” of the Forum is the “actual entities” (Whitehead’s language) it catalyzes and spawns. Whether the Forum itself has any sort of destiny in the world - or if it turns out to be rather ephemeral - is to me not a very central concern. The “actual entities” that engage me more are people, communities, organizations, larger social formations, and the historical process writ large. If the Forum pulls things together for entities that concern me, then it serves its function well!
Whitehead’s process philosophy is very relational and to me online Forums (not just this one) are seed beds of relationship. It’s like a greenhouse. The goal of the greenhouse is not to optimize greenhouse-ness. It’s to shelter and nurture plants. Unless the greenhouse ships new seedlings out in every season, it’s not really serving its function. Next year’s seedlings will look quite a bit like last year’s seedlings, however. so a year-over-year comparison of what’s happening in the greenhouse itself will not necessarily suggest “progress”. The progression is in the larger process.
I recall the thread started after exchanges with someone who was so fundamentally rooted in their position that absolute belief in its veracity offered no chance for any real debate. I remember feeling exhaustion and also internal conflict, for the ‘solution’ of exclusion never sits well, and likewise that person introduced me to the group.
We live in a world where regard for diverse perspectives and ideas is narrowing - a world of extremes. Ironically this reflects what is happening in nature, too - we may be more interconnected on a deeper level than first seen. The beauty of Whitehead’s process philosophy is it creates an arena of interconnection, rather than that of static absolutes. I like the greenhouse analogy, though can’t help wondering: what happens if one or several plants dominate? How do we ‘contain’ them - if not opting for exclusion? I believe that question was never really answered.
The other day I had to pull a bunch of Himalayan blackberries out of the edges of Greenhouse #4. In our area, Himalayan blackberries are a very invasive species that will gladly crowd out everything if given half a chance! Anyway, greenhouses - like online fora - need tending. Entropy inevitably returns everything to chaos, in the absence of energy expended to create order. So greenhouses without workers to tend them become ruins with the most aggressive species eventually taking over. The metaphor is still working quite well from my POV!
