Deeper into AI

In which we continue feeling our way through an A.I.-assisted post...

Happy New Year!!

I predict that this is the year when A.I. changes everything, whether we realize it or not. In that spirit, let’s continue to explore using A.I. to help write this series.

Idea Generation:

In the last post, we used Bing’s free chat tool to create an illustration and to make lists of life stages and professions to be used later. This exercise sparked a series of ideas, showcasing one of A.I.'s most valuable applications: the ability to ignite creative thoughts for further exploration.

Here’s my next prompt: “Please create a list of "passages" a person might expect to go through during their life. Examples include illnesses, financial troubles, loss of friends and family, or positive changes such as weddings, births, graduation, etc.”

…and a sampling of results:

• Birth of a sibling

• Starting school

• Making a best friend

• First love

• Graduating high school

• Leaving home

• Starting university or a job

Then, I edited my earlier prompt to incorporate these “passages” into this work and generate some more rules of thumb to work with. Here’s a sampling of the first few:

Birth of a sibling:

• Embrace change: Welcome new family dynamics with an open heart.

• Share and teach: Use this opportunity to teach and learn from each other.

• Patience is a virtue: Adjusting to a new sibling takes time for everyone.

Starting school:

• Lifelong learning: Approach education as a continuous journey.

• Friendship fosters growth: Cultivate positive relationships.

• Resilience overcomes obstacles: Persist through challenges.

While these rules are very generic and lack specific actionability, they serve as a foundation. The next step is to refine them into a set of ‘keepers’ that meet our established criteria.

What criteria should I use? Now that I have a substantial list of rules, it’s time to winnow them down to a useful set. So, here’s the next prompt I’ll try:

“Using the rules of thumb created from earlier in this thread, and generally accepted rules in society such as the golden rule, please use your best available resources to rewrite each rule based on the following criteria: 1. Each rule must be actionable and stated in a way that provides an example action step a person can take to implement it. 2. Each rule must increase the odds of a good outcome, in other words, it must deliver results better than random chance. 3. The "rules of thumb" should be concise, memorable, and clearly express the rule. Step through each rule as it is rewritten and score the result on a scale from 1 to 10 on the following parameters: 1. General usefulness (can it be helpful in many domains?) 2. How much does it contribute to the thriving of the individual? 3. How much would it contribute to society as a whole? Once the entire list of rules is created, provide a summary listing of scores and top ten rules.”

The initial results were somewhat unexpected:

Curiosity is key (Childhood)

• Actionable Rule: Ask “why” five times when encountering new information to deepen understanding.

• Scores: General usefulness: 8, Individual thriving: 9, Societal contribution: 7

1. Play is learning (Early Childhood)

   • Actionable Rule: Engage in one new playful activity each week to discover new interests and skills.

It turned out that the AI only considered the most recent list of rules it had created. It seems I waited too long to create the next prompt and found myself in a new session that had no way to refer back to my previous rules. This illustrates one of the problems with today's generation of AI. Each session is independent of all other sessions. There is no "memory" kept between sessions.

Also, the input limits of 4,000 tokens (roughly equivalent to words) create a serious limitation in the usefulness of today's AI.

So, I’m now going to try copying from my notes into the AI chat window with all of those earlier rules. This could be a problem since the AI is limited in the number of inputs it can accept. So, let’s try it and see what happens.

The first attempt will be to see if I can copy the entirety of the existing rules into the chat…

Success! Here’s the first part of the prompt. It took almost all of the 4000 tokens, so this is not the best approach when you need to cite larger sources, but it worked for this instance.

“Please refer to the information below as source material to rewrite and score as you did in the examples above: • Child • Teen • Young Adult • Marriage • Career • Parenthood • Middle Age • Golden Years Here is a list of 30 professions across various categories: 1. Project Manager”

The resulting list was quite long so I’ll show just a couple of examples from it:

Golden Years:

• Stay Active: Engage in a physical or social activity several times a week.

• Life Reflection: Write down or share a life story or lesson learned with someone younger once a month.

• Adapt and Learn: Try a new activity or learn something new every few months to stay mentally active.

Now, let’s move on to the professions:

Project Manager:

• Goal Setting: Start each project by clearly defining and communicating its goals.

• Risk Assessment: Conduct a risk assessment at the beginning of each project phase.

• Stakeholder Updates: Provide regular project updates to all stakeholders.

Next Steps

I have a couple more things to accomplish. One is to have the AI review this post and suggest edits. [I have incorporated those edits already] Finally, I’ll ask it to produce some visuals to accompany the story.

My first two attempts resulted in this message:

Oops! Try another prompt

Looks like there are some words that may be automatically blocked at this time. Sometimes even safe content can be blocked by mistake. Check our content policy to see how you can improve your prompt.

But I finally got some nice illustrations and chose this one.

That’s it for today’s post. See you next week with another mental excursion.