Awe Will! I'm so impressed and so jealous. All I can do is shake my head at your skills and passion for robots. I am truly looking forward to seeing your journey and all those robots lined up. Do you have a list of the ones you want to print and display?

Word of the day "Narny". LOL.

Hahah. Yeah I have alot going on in the coming months. I have several more channels I'm starting as well as updates to Reel Robots. I've ordered more 3d printers to start printing all the parts for the robots. I will be jumping all around from bot to bot while parts are printing. I'm also exploring replacing servos with BLDC motors and controllers through SimplyFOC. This will allow quiet operation and so I can go bigger with robots. The list of Movie Robots (not necessarily limited too or in this order) are:

1. B9 (video exploration of designing and animating)
2. Robby the Robot (same exploring)
3. K2SO (building, animating and exploring AI cloned voices for custom communication)
4. C-3PO (same as above)
5. R2-D2 (2-3-2 construction and animation) Then a slew of personal designs, (revisiting the spider bot, Alan and Alena as courseware for robotics) and many many more.

Another drop, This one resolves an issue with PI and its annoying tone between latency using an extension for Chrome called SAY,PI.

Wow. That was impressive and a bit intimidatingly. That robot head is obviously way smarter then me. Is there any way to dumb it down? Kinda to the level of the bar fly you meet on a Friday night out? I wouldn't feel so far out of my league then. LOL

It really sounds human. The voice cadence is amazingly human and it sounds like it's actually thinking with a brain. I think we have finally arrived at that point where we now have human like talking and reasoning computers like the science fiction computers of shows and movies like Star Trek and 2001 Space Odyssey. "Hal, open the pod bay doors". "Dave, I'm afraid I can't do that".

Hahahaah I felt the same way! She was keeping me on my toes! We’ve come a long way from our early days of canned responses with our AIML. It feels like a smooth flowing conversation with her asking me as many questions as I asked her. This is certainly the companion I was thinking of 10 years ago when I started building these robots. ( yup 10 years ago). AGI is right around the corner, then HAL will lock us all out of the bay doors!

I saw that and was hoping you were gonna post it, otherwise I was going to. I was one of your likes! It was a deep convo but pretty exciting to watch. I'm not sure if you have seen our latest open ai gpt robot skill update - but it can now generate code in it's responses that can be executed. It had me wonder if there's a way to merge the conversation that you've having into the open ai gpt robot skill for custom real-time programming of the ai to move itself.

You see, the way it works is by providing examples of how the robot can program itself. You can specify parameters it can change, and any examples. Code is taught to llm's so they understand it quite well. It can check variables and make decisions based on conditions, and create it's own loops and movements.

It's too bad many of these chatbot type programs that you're using are painting themselves in the corner by not allowing api's for this sort of thing.

I saw the open ai API is allowing audio to be transcribed and processed - but the challenge is still knowing when to listen to the audio etc... I'm sure we'll figure something out soon for that. But you won't have the speedy responses with open ai as you get with that llm. Mainly because the open ai gpt is so stuffed full of knowledge.

I did see when that skill (Chat GPT) came out but haven't had any time to check it out. I'll have a look today. Inception, the company that created PI, uses some of Chat GPT functions:

Quote:

Say, Pi uses OpenAI's Whisper ASR to transcribe spoken audio into text, and ChatGPT APIs to further enhance accuracy. All of which results in class-leading accuracy, with word error rates as low as 2%, much better than Inflection's native Call Pi STT feature.

Yeah Latency is the biggest hurdle to overcome. To accept the conversation as natural free flowing and organic, there must be the same latency there is in a regular dialog between two people.

• Not even in the same class as fxrtst!  But... @DJ,  I am going to spin up my C3PO head again with Open AI.  I had it working 6 months ago with the API.  Any advice on what path to take to get the best outcome?