Armed with a Pi-style SBC and your AI library of choice, I bet you could get pretty far implementing some toy-like behavior. Bonus points if you use Whisper for speech-to-text, and double brownie points if you can get an AI voice to read the generation back.
prophet, python talk to chatgpt [2]
call in to radio & get an on-the-air answer from chatgpt [3]
looks like text to image video of chatgpt answer generated in near real time might be more of a challenge.
aka degzo[4], but frame by frame 'animated movie' of the answer.
complete hardware/software project : https://www.hackster.io/nickbild/voicegpt-f88f8f
perhaps the ultimate in video compression -- convert video images to text, send via morse code and convert back to video without lossing video quality. [5] aka cons it in, cdr it out
2021 the year of the 21st century diy replicator -- voice to chatgtp to image to gcode to 3d printer.
wonder how long it'll take for chatgp image to 3d printer item to get things right 100% of the time without human intervention.[8] [6] [7]
[1] : https://github.com/C-Nedelcu/talk-to-chatgpt
[2] : https://www.youtube.com/watch?v=K_NRDg5Egrc
[3] : https://hackaday.com/2023/01/28/the-voice-of-chatgpt-is-now-...
[4] : https://dezgo.com/
[5] : https://www.producthunt.com/products/promptheus-talk-to-chat...
[6] : https://forum.duet3d.com/topic/31295/chatgpt-gcode-generator...
[7] : https://www.geeky-gadgets.com/cnc-pen-plotter-09-02-2023/
[8] : https://medium.com/@nilsbongarts/chatgpt-the-easy-way-to-fin...