Changes to My Blogging: AI

I'm going to start using generative AI (what people call "chatbot" technology) to help me create draft blog postings.    In all cases, I will review and edit every word to ensure three things:

• Every posting has the right tone.
• Every piece of information in the posting is accurate.
• Every posting conveys the truth, as I understand it, in terms of importance, future directions, risks, unknowns, and so on.  I check that the things implied the posting implies are true, are in fact, true.

This blog has never accepted as sufficient mere facts-and-figures accuracy (what the media often calls "fact checking").  In addition to checking facts, I have always spent a lot of time checking that the tone, implications and views conveyed in this blog are also the truth, as far as I know it.  That will continue, even as I use AI to help write initial drafts of the blog.

I'm starting to use AI for three related reasons:  First, blogs take a lot of time and I hope that using generative AI will allow me to create blogs of the same quality in less time.  I'm not going to save time in the editing or reviewing part of writing blogs, but crafting the English takes me a lot of time and I hope AI can speed the process. Second, I hope that I can use this newly available time to either create more blogs or create more complex blogs.   I want to do more analysis and historical perspective type blogs, rather than just writing all the time to keep up with the clinical trial results.  I feel that is sort of the news equivalent of "treading water", and what I want is to get ahead of the news and use the news to write more inciteful posts.  Third, it would be nice to get news about clinical trials out quicker so you can read about it sooner after the results are announced.

In terms of "how will I use AI", I'm not going to use it in a simple minded or unthinking way! I am going to experiment with several different methods and expect to end up using a combination of these techniques, and incrementally improving them over time.  

• For example, one simple technique would be to have many different chatbots respond to the same question, and then I, as a human, merge the best parts (or the best language) from each of their responses together to create a posting.   
• Another technique I will try is asking very specific, very focused questions of AI in order to generate parts of the blog posting.  Then (as a human) merging those parts together into a posting.
• I will also experiment with creating very complex detailed questions for the AI in order to generate blog postings all at once.  (This is called "prompt engineering".)

The future is the undiscovered territory, and I am going exploring.

Joshua Levy

http://cureresearch4type1diabetes.blogspot.com

publicjoshualevy at gmail dot com

All the views expressed here are those of Joshua Levy, and nothing here is official BreakthroughT1D or JDCA news, views, policies or opinions.  I sometimes use generative AI ("chatbots") to generate draft blogs, parts of blogs, or drafter alternate wordings for these blogs.  I always review every part of every published blog to ensure that it is saying what I want, in the tone that I want, truthfully, and accurately.  My kid has type-1 diabetes and has participated in clinical trials, which might be discussed here.  I am obese and right on the boarder of T2D and therefore may be taking drugs for those conditions.  My blog contains a more complete non-conflict of interest statement. Thanks to everyone who helps with the blog!