AI Powered Bots
I'm tinkering with several AI powered bots that you might enjoy playing with or find useful, just install Telegram and click below. If they are unresponsive you might need to try later; they are all housed in a 2007 Macbook so it only takes a few concurrent users to crash them.
| Model | Avatar | Features |
|---|---|---|
 |
Protocol droid. Type /protocol to activate different modes, like a mechanic to solve your car issues or a comedian to make them more palatable. You can send voice notes and hear its responses if you want to talk rather than write/read. Connected to OpenAI's GPT4 and Google's Text-to-Speech via API. |
|
 |
Illustrator. Generates images based on what you write or the voice notes you send. Credentials include lots of half baked bot avatars. Hopes to learn to communicate with Midjourney v5 when it grows up. Connected to OpenAI's DALL-E and Google's Text-to-Speech via API. |
|
 |
Story-teller. Write or send a voice note with the first sentence and it will provide an illustrated story with several chapters that it can read for you. Connected to OpenAI's GPT4 and DALL-E, and Google's Text-to-Speech, via API. |
|
 |
Voice transformer. A bot that changes to different voices text and voice notes you send. Connected to ElevenLab's Synthesis and OpenAI's Whisper via API. |
|
 |
Image-eccehomo. A bot that butchers images you send, unless they are very basic in which case it returns an image which is somehow similar. Connected OpenAI's DALL-E and Google's Text-to-Speech via API |
|
 |
Image-scribe. Extracts text from images, including photos you take directly via the Telegram app. Connected to Google's Tesseract and Text-to-Speech APIs. |
|
 |
Audio-scribe. Transcribes audio notes and mp3 files in most languages. Connected to the OpenAI's Whisper API. |