In lieu of the current lip-sync mechanism, do you have any plans to try giving the AI model more verbs to work with the motors (within safe thresholds), and then encouraging it to do more than just lip-sync? Lip-sync should clearly be the default for normal verbal responses, but you could also encourage responses like jaw-drop reactions, or patterns (would be funny to see it using its tail to tap out "shave and a haircut" as a bit).
It will think (generate tokens) throughout the day, then convert that to voice and save it and will keep generating so on. Every time it is poked, it will play one of the saved recordings.
Replace the bedrock setup for a gemma-4 E2B one.
EDIT: Looks like modern versions of Billy have a motion sensor too--that could be fun to hack as well.
I don't know, some of the news today is so preposterous that it could only be delivered by a BillyBass.
1) Use a leaded solder with flux mixed in like the Multicore 502 [0]
2) Use more flux.
3) Use little alligator clips on adjustable arms to hold the stuff together for you so you can keep adding… yep… more flux.
4) Get some braided copper solder wick with flux in it for when you’ve finally added too much solder but it still looks terrible and you just wanna start over. After wicking away all the solder, clean up the wires with some more flux.
Oh also the flux is more toxic than the lead. Get some good ventilation so you can use more flux.
0: https://www.newark.com/multicore-loctite/c-502-99c-5c-0-7mm-...