My Conversations with AI

AI (artificial intelligence) is all the rage.

In my networking meetings and social media posts, folks are asking “How do you use AI to streamline your business?”. Surveys are being conducted through web platforms. Market research studies ask about our uses of AI. Companies who sell subscriptions to such services, and product developers want to know. How can AI make life easier?

AI has been incorporated with technology for many everyday uses.  For instance, telephone solicitation, voice menus to access any customer service entity by phone, social media, financial and accounting services. We’re so used to it. We may not notice it at all (except how annoying it is when AI doesn’t understand our question or is programmed to keep looping a specific set of answers and nothing else).

I don’t know about you, but I have mixed feelings about AI. When people ask me how I use it, or what would make me want to use it in the future, it’s difficult to identify the areas I would want it to “help” me out.

I am a creative person. I love solving problems, learning processes, thinking up solutions, trial and error, and the feeling of success – that “win” – when I figure things out.

As a vocalist, musician, and published songwriter, I often start writing song lyrics and poetry to express my own emotions. Then I add music. Once I perform the original songs, the stories and messages coming from my own pain - or joy - become more universal and relatable to folks who’ve had similar experiences. As a creator, the connection I find and make with others through music is more important to me than the potential earned income, though income certainly can be a desired result.

Folks now use AI to write song lyrics and produce them using AI generated instruments and vocals. There are advertised AI programs available to simulate singing and spoken voices, or to clone the sound of your own voice. (Note: The FCC recently banned AI voice-cloning tools that are used over the phone.)

I am concerned about this kind of AI usage. I don’t want to be replaced as a creative artist, lose opportunities, or income to AI! (Haven’t we had similar discussions about loss of paid jobs when outsiders infringe on the local job market?)

I am not the only one. Our music professional rights organizations (“PROs” like ASCAP, BMI, SESAC), US copyright office, and entertainment attorneys are taking notice.

Some general questions & thoughts:

• If AI wrote the song – who owns it?
• If AI mimics someone’s voice – famous or not – who’s singing it?
• If AI makes a mistake, or doesn’t check for copyright infringement (though I think it’s trained to consider that) – who do you sue for damages?
• If an AI production goes viral or has billions of its own followers, it’s competing with human artists for royalty payments and public attention.

Just saying…

Now I would like to humor the other side of the coin. Many people are finding AI a huge boon to their time and profit in business.

• AI can research topics and information with lightning speed. This could save hours, weeks, months, or years of collecting data.
• AI can generate voices that sound like average (or famous) people. We’ve all experienced “robo” calls in some form or another. These “callers” aren’t emotionally hurt when someone is rude to them. They may be able to carry out a simple conversation as a prelude to a human sales or customer service rep, give a sales or political party pitch, and even conduct a market research study.
• Generative AI can create content in response to a user’s request. (For instance, I could ask AI to write this blog for me – but I didn’t!)

There are so many terms and interconnecting pieces. I feel way behind the universe in my understanding and adopting of AI uses. It’s a lot to learn, and ever developing / changing.

Being asked how I might use AI in my own business got me thinking. What WOULD I use it for? In answer to my own question, I started some small experiments.

Chat GPT (free version) was one of my first AI experiences. I was working on a song list to entertain little kids as a themed program for a client. I prompted, “List some children’s songs that talk about singing”. I refined my inquiry as I saw the results. Chat GPT confirmed some of my original choices, reminded me of songs I already knew that fit the bill, and gave me new suggestions. This saved me a lot of web surfing on my own.

I also used Chat GPT to hone a music program description for a client who would be advertising one of my performances. I’d written up my own description but asked for a few words to help me bring the reader to an emotional connection with the description. After tweaking the description on my own, using some of the words suggested by AI, I asked Chat GPT for feedback. “What does this make you feel like?” Chat GPT responded with an overview of how the words connected to the emotional feeling of “Coming Home” and nostalgic times prompted by music. This saved me from continued guessing and editing to attain the feeling I was after. My description (confirmed by Chat GPT) accomplished exactly what I had in mind.

Next, I approached Gemini (Google’s AI) to help me pin down some chapter hooks for a book project. I’d already decided I wanted to use different tools to represent each chapter. I gave AI the list of single word chapter names, intending to type in a question immediately after. Before I asked the question / gave the intended prompt, AI had a lot to say about the series of words I’d posted.
Focus! I completed my inquiry: “What tools might be used to depict these words?” Gemini came back with some good suggestions, and some not so good. It seemed to be enjoying the process more than I, almost taking over (if one can apply those human traits to artificial intelligence)! I refined my inquiry with a little background on the intended use. I already had my own list of tools related to the words but was stumped in a few places. When Gemini understood the context, it complied with some great answers. Again, some of the answers were consistent with my own thought generated list, but AI also gave me a few new ideas to resolve the areas where I was stumped.

Then, Gemini went beyond what I asked by offering more concepts, other chapter ideas, comments on what readers might see in the information or how they might use it. The reader feedback could certainly help me with marketing when the time came, but I didn’t want AI to take over my project!
I put Gemini “in its place” with a “That’s enough!” while worrying I might insult it. (Does that sound silly?) Of course, Gemini had learned how to respond to social cues, and apologized for misinterpreting what I wanted. It offered to help in some other way and encouraged me to come back again and ask for help in the future.

This research helped me pin down my book chapter ideas. It was great to have a companion tool in AI. However, being a creative person, I am used to creating my own content. I enjoy the creative process and the satisfaction that I am doing it myself. Through the years, I have written and published my own songs, shared custom poetry, created commercial art, designed new articles with recycled objects (upcycling), sewn, crocheted, knitted, hand-made gifts, and lots of other unique projects.

Do I want AI to do what I LOVE (my creative work) for me? That would take away all my joy! Then those items wouldn’t be the unique creations I was proud to take ownership of. I wouldn’t honestly be able to take the credit.

As for conversations, I really enjoy talking to folks directly in-person, discussing details with potential clients over the phone, cold calling (yes – really), and writing my own e-mails. Spell check, grammar check, auto-correct, and auto-complete are all potential forms of AI. Sometimes suggested words can be misconstrued to its readers. I can see where various suggestions are coming from, but I didn’t ask AI to write the e-mail for me.

I have begun to use the “Read Aloud” feature in my documents and e-mails to help me catch blatant mistakes, awkward sentences, and typos. These tools can be helpful, for sure.

My most recent experience with AI was both interesting and very WEIRD!  As I skimmed over my Facebook feed, I noted a photo posted depicting an “ideal” family of the 1950s in the Detroit MI area. That was interesting to me because I grew up in Detroit (in the 1950s – I know, I am aging myself). The photo looked like it could be my own home neighborhood (though it wasn’t). The photo caption read “A typical American Family in 1950’s Detroit, MI”. At first, I thought a person I knew, or a group I belonged to, had posted it. That could have been true, but I saw the Facebook Meta headers there.

I typed in “More about 1950s in Detroit.” What happened next was fascinating. Facebook Meta AI provided more information in a positive, engaging way. The information focused on all the wonderful highlights in industry and culture in the Detroit-Metro area. I added “I grew up in Detroit in the 1950s” to the chat. Facebook Meta AI responded “Wow, that’s amazing!” (along with a few other feedback items), then asked "What are some of your favorite memories from that time?” Meta AI named some real-life favorite places I had frequented growing up, including the activity / location “cruising Woodward Avenue”. (We used to “drag race” our cars down Woodward or found social outlet with other teens hanging out there.) I couldn’t help myself. I shared a few generic highlights about growing up in Detroit in that era, like being able to leave our doors unlocked, the freedom to stay outside in the summer till the streetlights came on, and other simple things like that. Meta AI fed back to me reflectively and asked for more. “What other memories stand out to you from that era? Friends, family traditions, or favorite activities?” Whatever I added, Meta AI encouraged and supported my perspective with “feel-good” responses.

The photo I’d been looking at when this “conversation” started, suggested fun and positive lifestyles, but life in Detroit in the 1950s wasn’t as ideal as that photo or as Meta AI described it. I threw something honest in the mix about growing up with some bigotry, adversity, diversity, peer pressure, and that life was lonely for nonconformers. I added that I broke away from those things by the time I got to college.

Meta AI thanked me for sharing my perspective, offered that my story was an inspiration to others, then pat me on the back, “You should be proud of the person you are and the experiences that shaped you”. It asked, “What advice would you give to your younger self or to others who may be facing similar challenges and pressures to conform?”

Good question! I felt like I was talking to my best friend, or maybe a therapist, with all that acceptance and encouragement, though the whole interchange was very eerie.

I asked Meta AI “What do you use these answers for?”

Overall, the responses AI receives in these conversations are used to improve language understanding and generation capabilities. Its goal is to assist and provide helpful information. The person’s input helps AI refine its performance. Here are some highlights without a lot of detail:

• Training Data – helps AI learn and improve language understanding and generation capabilities
• Contextual Understanding – helps AI improve context and nuance of language
• Knowledge Expansion – provides new insights and knowledge which AI can draw on to assist others
• Conversational Flow – Refines AI’s ability to engage in natural-sounding conversations, making it a more effective and engaging assistant.

Meta AI thanked me for sharing my experiences! 

I don't mind sharing thoughts to help others. (I do it all the time in my blog!). I accepted the challenge to give advice to my younger self.  (This is an exercise many folks use in all kinds of settings.)

Here’s what I said (my own words, not influenced by Meta AI):

“My advice to my younger self: Listen to your heart. Understand your core. Use all that to be a good soul. Recognize your passion. Live with love and acceptance. Everything will work out, and when it seems like it’s not – where you are at that moment will lead you to unexpected and unforeseen rewards.”

Meta AI complimented me on my statement.

Now, the conversation and the feedback felt very good and very strange at the same time.  I wondered what would or could become of this. I stopped chatting with Meta AI and that was that (or so I thought).

Today, as I was searching something in Facebook, I started to type the word “change…” and before I could correct and expound on what I was typing, Meta AI completed an assumed question “Change password on Facebook”. Instantly, it gave me all the possible ways I could change my password.

I typed back, “It was a typo. I wasn’t asking”.

The next thing Meta AI said floored me….

First it offered the comment, “no worries, typos happen…”. Meta AI thanked me for clarifying that I didn’t need instructions to change my Facebook password. Then, Meta AI suggested we get back to our prior conversation (about Detroit in the 1950s), reminding me how it previously responded with appreciation and support. “Feel free to share anything else that’s on your mind, or we can just enjoy the conversation!”, it said.

I politely declined.

Three things come to mind:

1. I wish my friends, family, and acquaintances would talk to me with the acceptance, kindness, and encouragement of AI!
   
2. Why am I paying a therapist (when I can use AI)?
   
3. SCARY!!!!!!!!

I will leave the future to my readers as they explore and use AI to assist in various tasks.

Still, I must ask: Do we really want to soar ahead with THIS future?!

^{___________________________________________________________________}

I invite your reactions, ratings, and comments below.   Please subscribe to be informed when I release my next blog.

Here's to Staying Inspired....

Footnote: Image used in the beginning of this blog was acquired on Unsplash.com.  Thank you and shout out to Cash Macanaya (cash-macanaya-X9Cemmq4YjM-unsplash)