Artificial Intelligence Tracker

Table of Contents

• AI Models
• AI Tools
• AI Model Costs
• Tool deep dive
  • Coding
  • Presentation
  • App Deployment
  • App/Website Designs
  • Search
  • Voice to Text
  • Text to Voice
  • Document Processing
  • Voice Conversations

AI Models

There are tons of AI models out there. Most are propriety. Some are open-source. Prices are being driven far lower now. The Model Comparison: https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard

AI Tools

Model Costs

Last updated: March 28, 2024

Text to Voice

• Open AI: https://platform.openai.com/docs/guides/text-to-speech
• Eleven Labs: https://elevenlabs.io/

Voice Conversations

• https://demo.hume.ai/: haven't tried it yet.
• Talkio AI: pretty bad

Voice to Text

1. Agua Voice - $10/month (comes with an unlimited amount of voice from there. It's extremely good. It's by far the most intelligent I've ever seen. I think it's at least a 10x improvement over the other producers in the space.)
2. Otter AI is an app I've used for many, many years. It was pretty good, but it had a lot of errors. And unlike Agua Voice, I can't edit as I'm speaking. So it left me with notes that were hard to immediately put into blogs. It also isn't nearly as powerful from a context basis as a voice. I can't do things like tell it to translate entire thing into Spanish. I can't tell it to format something as markdown. It's just very hard to use OtterAI for kind of more powerful use cases.

Document Processing

These are a few things that I presented at the AI Whispers Meetup in Buenos Aires on March 27th, 2024. Really interesting stuff. Tons of really interesting uses out there.

1. Translating text

   And so what I did for this use case was I wrote a program and the repo is publicly available at GitHub.com/AlephDAO/translation-app. And what this does is it uses the OpenAI API to convert a text file into a translation. You give it a text file, a source language, and a target language. It'll translate it. It costs about less than 10 cents per 100 pages. It processes within a few minutes. And then you have your text file. It's pretty much perfect, even using chat 3.5, it works basically perfectly. Really, really good. You can see the code and kind of the readme file goes into more about that.

2. Using AI for PDF cleaning

   All of the tooling around converting from a PDF into a text file or some other format just sucks, unless you pay tons of money for a top Adobe product. I figured out a solution that I've already published on my GitHub github.com/alephdao that will allow you to easily clean a PDF into a text file, specifically a markdown format. And so how does this PDF cleaner work?

   What it does is it's a Python script that uses PDF plumber to convert the PDF into Markdown. And then from there, I feed that PDF into my LLM, my OpenAI API. And what that lets me do is then I have I'm able to clean up the spelling errors, the formatting errors. And it works really well now for eBooks. So you can try it out yourself. It's free to use. And I'm going to be testing it out on more difficult use cases soon.

3. Converting Between Different File Types and Publishing Content

   In this new section, what you can do is once you have it in text file formats, you can convert between that really easily. You can convert between a text file to a markdown file. You can convert from markdown and publish it as an ePub. So e-book. And then you can immediately view that in any of your e-reader apps. So it's really, really powerful. AIs are really good at taking text.

Return to Aleph DAO home page