This video's sorta all over the place, but I like it anyway. Don't get Terminated, squidlings! kaggle.com/datasets/NUFORC/ufo-sightings github.com/TedTinker/ufo_map arxiv.org/abs/1703.04908 patreon.com/thinkstr
Пікірлер: 6
@josephvanname33778 ай бұрын
Another video made for chlurmcks who know absolutely nothing about reversible computation.
@Thinkstr8 ай бұрын
Haha, we're all chlurmcks sometimes
@josephvanname33778 ай бұрын
@@Thinkstr Reversible computation is the future.
@VK-sz4it11 ай бұрын
Haha, we had similar idea at the same time. I have posted in another social media thoughts on how chatGPT can look so sentient and smart while having no way to test real world. You used Plato's ideas and I used idea of Lacan that what people call reality is not reality per se. Lacan argued that we live in 3 planes: the real (actual reality), the imaginary (representation of reality without using symbols and language) and the symbolic (language, symbolls, etc.). And what people mundanly percieve as reality - is born from symbolic space (My understanding of what he meant is the real - is kinda directly immesurable in similar sense that it is beyond Markov blanket - basically actual real world, the imaginary and the symbolic are representations of this real world. Imaginary - less abstract, symbolic - more abstract. ChatGPT is, in some sense, creature that inhabits only symbolic space - and it seems quite smart! It makes you wonder that Lacan was right here (and probably Plato too). In other words, makes you wonder, what percent of what constitues percieved reality is born from laws of language alone? Another interesting thing, Lacan said was that by developing skills in symbolic space, person looses big part of who he/she is - but that person has no good choice. In order to move forward one has to define self in more restrictive space - where it has less degrees of freedom - but at least it can compete and do something in world. PS.: I posted those ideas 8 days ago and watched your video just now.
@Thinkstr11 ай бұрын
I like those three planes; it reminds me of reinforcement learning agents having observations of their environment, but also having inner-states or hidden-states representing their embodied world-view. And if the agents can communicate, they can share information in very compressed ways! Thanks for watching again.