Well, my code I made is essentially GPT, half coded, and what is coded in my AI so far is in a slightly different way, and to get GPT results would require a few more things, which is very tedious to code in, yet small in my mind of a challenge, so I already know it definitely would not take 6000 lines of C++, it can be done in maybe ~400 lines of Python. C++ may be ~600 lines.
You can see in the code of GPT that the code is not 300/1000 lines, those 300/1000 lines are calling C++ code, like a controlling controlling a puppet, none of those 300 lines in Python are the actual algorithms. It imports Pytorch/etc to do so. There is lots of spaces and comments. The many files each containing code require code to call each other, which could be eliminated if were in 1 file.cpp. And the functions are probably bloated and could be made much smaller, they often make classes and indents and stuff that are not any helpful than a compressed form that runs as fast/ uses same RAM. My theory proves how Prediction is done, so I know how GPT/ my AI works, especially how mine works. Getting those mechanisms working is a LOT harder, even though I know exactly how they work and how to code them for my next steps. Still takes months. Besides Prediction, there is another part to AGI, this sort of changes what Prediction means, but maybe I'll tell yous another time more about it. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tc124b3d00b83e897-M75611eaba912ee2fde8c213b Delivery options: https://agi.topicbox.com/groups/agi/subscription
