From:
      David Lynch <dnl1960 at yahoo.com>
      To: Jon Oleksiuk <jonoleksiuk at gmail.com>
      Sent: Saturday, August 31, 2024 at 07:09:24 PM EDT
      Subject: Your The Trinity YouTube Video
      
      Jon,
      
      Regarding your Ai debating video series.
      
      "Jesus is God? AI debates Christian vs. Muslim"
      https://youtu.be/GOOgHN24WcQ
      
      I really like your concept, and I would like to create a similar system to
      flush out a theory that I have been working on for over two decades.
      
      I would require a ternary debate system. Science~Philosophy~Theology. 
      
      An objective viewpoint, a subjective viewpoint, and an imaginative
      viewpoint. 
      
      Your method seems to depend on Ai LLMs that are heavily biased by their
      foundation data set and their alignment.
      
      The Muslim Christian debate, points are given for quoting scripture like
      the one that uses the shedding of blood.
      
      Just because someone is said that blood must be shed does not mean that
      shedding blood it the only method. By giving high points for just quoting
      scripture is inaccurate. Thus the results are skewed.
      
      By using LLMs that are trained in scripture, the debate will always be a
      regurgitation of the LLM’s training data set.
      
      In another one of your videos, what you are exposing is the internal bias
      of Ai systems. 
      
      The one outlier was ChatGPT 4o.
      
      I want to build my ternary system, but I do not have the money to found,
      train, and align three LLMs.
      
      I would like to try building my ternary system using current LLMs, but I
      am not sure how you constructed your binary debate system. 
      
      Would you be so kind as to share the method you used to build your test
      system?
      
      If you want to understand my ternary KnoWellian Universe Theory, you can
      feed my “Anthology” into Gemini 1.5 Pro and ask Gemini questions about me
      and my KnoWell.
      http://lynchphoto.com/anthology
      
      Thank you,
      David Noel Lynch