Wittgenstein’s Legacy: The Principles of the Private Language Arguments (2018)

The universe, fortunately, has a plethora of similarities, such that some basic edge detection and a bit of neural net work in the retinal layers, combined with a bit of proprioception correlation, provides an enormous amount of 90%+ confidence-level shared realities, or at least a good enough fake. Just not 100%. GANS are doing a great job of showing us that not even all of that is required to begin the "faking-out" process.

But the illusion here is the same: that given this natural input we begin processing before birth, that these concepts extend to completely invented terms. Most folks never look there, they never wonder why a car is called a car, and there's no downside at all. It's a pernicious concept and a wickedly-difficult thing to eventually realize.

I don't see where we disagree. The only thing I'd add is that whether you have a completely private language or not, in terms of problem-solving/goal-seeking, is not important. For non-formal, non-tech things, using common words and gestures provides the quickest way forward. Once you start creating a self-consistent system of symbols representing state and behavior, though, you might actually be better off if everybody has completely different private languages. The illusion of common understanding where there is none is more dangerous than misunderstanding. There are no red lights or sirens that go off when human communication failures happen. It's all silence. It could be no other way.

I disagree. Tools are just based on what is possible and what seems to work. No deep understanding is necessary.

Text messages turned out to be a great tool, created more or less by accident and intended for a completely different purpose.

Twitter is something no-one asked for, but it changed the very landscape of mass communication. Its more like a tool that has shaped the world around it.

Wikipedia is a fantastic tool. It was essentially created on a hunch, that a crowd-sourced encyclopedia would work. Deep analysis of the fundamentals of knowledge and how people work together to shape it was not necessary for it to be built. It was more like 'this is possible, lets try it'.

The internet is basically layers of hacks, each bent to a different purpose from the original design. If you try to design something like the internet and make it philosophically pure, you end up with Project Xanadu. Which doesn't work.

Facebook's model of how human friendships work is laughably simplistic - binary on/off friend/unfriend. There's no insight there. If the Social Network is to be believed, Zuckerberg's main insight was to build something that would help harvard students get laid. But still, Facebook works as a tool. Google Plus tried to have a more accurate model of relationships, and made it too complicated. Perhaps insight is a disadvantage? A good tool should be about the possible.

I’ll let professional philosophers try to peel apart every nuance of the Investigations. It is a thoroughly enjoyable and approachable—if not chaotich—book for anyone used to reading academic papers in CS, math, or linguistics. The topic is deeply scrutinizing our everyday language, especially language about mental states. He is a very persuasive and poetic writer. Some excerpts:

- To utter a word is to strike a note on the keyboard of the imagination.

359. Could a machine think? ... Well is the human body to be considered such a machine? It surely comes as close as possible to being such a machine.

583. ... And the word “hope” refers to a phenomenon of human life. (A smiling mouth smiles only in a human face.)

So, one grain of sand is a heap?

If one cell is enough, than one atom is enough and so on. Then, nothing's really the same. But we use this notion because it's convenient, so we're back at the beginning.

"This has dramatic implications. For it immediately illuminates venerable puzzles concerning what non-human animals can think or experience, and whether they can reason or act for reasons. It sheds no less light on the question of whether machines can think, or whether we shall be able, sometime in the future, to make machines that can think. It is not surprising that Wittgenstein exclaimed ‘How much one must be able to do before one can be said to think’. It has equally dramatic consequences for theology, both for the belief that we survive, bodilessly, after death and for the idea that there can be a supreme thinking being that has no body."

Wittgentein's ideas have echoes and deep relationships with modern subjects like cognitive science and artificial intelligence (and sensitive ones, like religion and beliefs in the supernatural). In the case of AI, what is interesting is that his ideas do not discuss whether it's possible or not to create a thinking machine (i.e. technically or scientifically), but rather what it would really mean (conceptually).