@jpfraneto
another epic work session today
as i told you a few days ago, i am going to apply to the pump fun's hackathon with $anky. for this, i need to create a video. and i have been thinking how to create the video for the past days
so what i did today was to send some of my favorite teachings from my meditation teacher (8 selected youtube videos from the awaken the world youtube channel - about 15hr of videos) to claude and have it write a pipeline that transforms the core teachings behind those videos into a script that presents *anky* as a meme that wraps a spiritual practice
and then to generate prompts for images on the video every 4 seconds. scenes. and then on the pipeline i have two ai video models that will transform those images into videos. in paralell, to see the difference between them
and then the pipeline stitches everything together and transforms it into a 2:22 video "about anky and what it brings to the world"
after that process is done, with this final video i add TTS (text to speech) to have another ai read the script and add it as the background of the video
im ready to go home rn and leave it running on the office. i don't really have a clue about what will come out of it. but i do know something:
i will hit enter on the computer and "stuff will happen". and the output of that stuff will be two videos of 2:22
good? bad? slop? not? who cares
the important thing is knowing that this technology will just keep evolving. and that this will be the "first autonomous anky video ever"
and that sole thought brings chills to me
(this video is from one of the tests)