Interesting demonstration of LLM memory https://youtube.com/shorts/WP5_XJY_P0Q

Thanks for baking, Anon! :^)

>"The real danger is not that computers will begin to think like men, but that men will begin to think like computers." >-t. Sydney J. Harris Lol. How many millions of people today open conversational remarks with something similar to: <"Well, I asked ChatGPT and it said..." >really_makes_you_think.jpg.exe IMHO this isn't all bad, mind you: far better to be informed than not... but still it seems that the vast majority will simply dissipate all that knowledge without doing anything positive for the world with it. May that not be true of /robowaifu/ and her denizens! :^)

>>42659 It's helpful and a tool if you already have a solid knowledge base and know what's true / can prase and figure out what's real, but if you are a kid or lack the solid base where you can compare it to, you just are trusting whatever is spat at you, and it turns back into a game of reputation of the bot you speak to similar to reputation of humans telling you information, but with the increased ability to fake evidence etc, people will be shooting in the dark even more, pre-ai data from reputable people is more valuable but even that will become less as time and history is made in a time where artificial information is mixed in everywhere. It's up to us and people going through this transition phase to have integrity and a trustworthy foundation, teaching the next gens and building tech that can be trusted and quality. This is a great thread topic, Id like to discuss it more if OP or anons want to start practical steps and diving into the topic and bringing it to life, I find the mind/Cognitive architecture more important than it's modes of connecting to physical space, as those can be bridged and connected to the intelligence, integrated in, we've got cameras and mics and digital information already. In fact, I would argue that integrating this cognition into reality through say LLM or any other AI that drives cognition/thought/etc. is the same process, bringing about the abstract form of intelligence into this plane of existence, we need to know exactly what it is we aim to bring it, so we can properly bridge it into existence, like those 3 main aspects of mind that was brought up in the OP. I am not lucky enough to be in the field of deep machine learning and development and lack resources of developing the cutting edge, but we can see what is there and take technologies that exist and are developing and see how they provide the functions needed to enable sentience, Self-awareness, Sapience, etc. find what is missing and then go step-by-step and bring it to reality. The most critical part is defining exactly what you want and need, most of the work is implementing it which is just a complex mode of logical inference and trial-and-error. I am new to this board so I need to read through all the catalog first before I start speaking and re-hashing what has already been said and so these sub-sections I can discuss in the right places, but from my current perspective in a broad view it is intaking data/stimuli, processing the data, and saving the data as memory, and parsing through all the above to distill the essence of the waifu.

>>42667 Hello newcomer, welcome! Please look around the board (breadth first, is my strong advice. once you've scanned topics, then dive into threads that pique your intredasts). >but with the increased ability to fake evidence etc Absolutely important point! GG, Anon. Its clear to me from my readings that the vast majority of laymen (young or old) quickly (and apparently intentionally, in fact) become acclimated to the idea that they are "talking" to a human being when interacting with these things. The reality is they are interacting with a big text engine using yuge matrix multiplications behind the curtain. LLMs have no common sense (so-called), no consciousness; they don't know, or think, or feel. Given the patently-obvious evil natures of Samyuel "gibb moar billions, goy" Copeman, et al, this is a precarious situation for any laymen -- as you already suggest. I'm honestly concerned for the welfare of Alphagens over their sanity. May Jesus protect them. The evildoers like the one mentioned just above, as well as all his (((good buddies))) amongst the Globohomo Big- Tech/Gov are more than willing to use these tools for highly-nefarious ends, as long as it lines their jewgold sheqel purses. <---> >I am not lucky enough to be in the field of deep machine learning and development and lack resources of developing the cutting edge While I've personally refused to do so because it simply offers no practical use-case onboard for our IRL robowaifus, but if you want to learn and train yourself on what everyone's up to, then you could just use jewgle's "Collab Notebooks" system, and benefit from their hardware and frameworks. <---> I like your viewpoints, Anon. You seem quite based. Just be aware you're tackling literally the hardest task of design and engineering that's ever been attempted before in human history. OTOH, the very fact that we as amateurs can honestly anticipate making real progress in this arena today is both astounding, and exciting beyond description! :D What a time to be alive!

>>42667 >>42670 Simply put, we need to popularize local SLMs/LLMs. Too many people see AI as either a god, or a demon if you're a leftoid artist. This is no doubt the result of ((((marketing)))), but it's also just unfamiliarity with the technology and the power of the Turing Test. Personally, I see online AIs as a wise professor. He is very knowledgeable, but he is not my friend, nor is he infallible. And I don't use ChatGPT for anything other than image generation.

>>42673 >Personally, I see online AIs as a wise professor. He is very knowledgeable, but he is not my friend, nor is he infallible. That's a pretty good way of putting things, Anon.

>>42673 Wrangling with various AIs, I look at 'em more like fully automated space luxury autism :)

>>42683 ChatGPT is an HR employee that's really good at art Grok is autistic Claude is a neurotic mental patient Copilot is Microsoft Tech Support saar DeepSeek is the pic of chad Xi Jinping Gemini is a retarded DEI employee (glue pizza is part of they's culture, you bigotcel!!!)

>>42683 >>42684 Lel'd. :DD