Making A Difference Podcast

Hillarious Hallucinations

Long time readers will know that I've been leveraging Google NotebookLM for AI generated conversations of my posts.

For the most part, they are very good. There are always interesting ideas, observations, and sometimes fascinating twists on the writing. There are some hallicinations, but many that are knowable only to me. Listeners may not even recognize these as hallucinations.

But recently, many of my posts have become more analytic and problem solving in nature. Some very directive on "here are the equations and analysis we use."

And in those conversations, I notice the models really struggle. They tend to invent completely false concepts, ideas, acting randomly.

For a test, I wrote a very short, succinct post. No hidden message, no hidden pattern, just an idea. Here it is:

1+1= 2

2+2 =4

3x3=9

4x4=16

5-1=4

That's it, just 5 rows of math calculations (I know they are correct, I used my calculator.) I intended them to be nothing more than what you see.

Now, I will ask you to scroll to the bottom and listen to the 5 minute discussion and insights into these 5 rows of equations. They get into the questions of the history of numbers, whether there is another way to interpret 1+1, and all sorts of other "serious discussions."

All completely fabricated, nothing having to do with what I wrote. Any human observer looking at the same "post" might say, "Dave, you doing your homework online? Dave, did you have to use your fingers and toes to figure these out......"

In just this simple post, we see so much of the weaknesses inherent in these tools. They have no understanding of meaning. They have no understanding of intent. They have no reasoning power.

Then this causes me to reflect back on my "idea" posts. The posts where I'm talking about ideas, concepts, observations, sometimes provocation. Are the conversations being generated for these as flawed as the conversation 5 equations? Are the hallucinations as bad, is the construct as bad, but do I fail to recognize them because of the abstraction of the ideas.

We have come so far with these tools. But the more I use them, I realize how very far we have to go.

Enjoy this discussion of the meaning of 5 lines of arithmetic equations.  It's a little frightening.  At the same time, the stories and discussion they invent are hilarious when you realize the total absence of meaning in the discussion.

Unsmart, Stupid, Willfully Ignorant

Here's the AI generated discussion of this post.  I discovered a lot about AI in listening to this discussion.  Reflecting back on this article, I realize at least what we are currently seeing in AI is often Stupid.  They assemble words, based on tokens, on something of which they know nothing about.  And it doesn't recognize it is Stupid.  And we sometimes miss this, the responses we get "sound smart," but they are actually Stupid, stumbling on truth accidentally.

There is so much wrong in this discussion, but it actually vividly illustrates much of what I am trying to discuss in this article.

The one redeeming thing is, at least for the present, these tools have no reasoning capability.  So it cannot be Willfully Ignorant.

If only for the irony and subtle comedy, enjoy this discussion.

Here is the link to the original post:  https://partnersinexcellenceblog.com/unsmart-stupid-willfully-ignorant/

The Sales Process And Contextual Flexibility

This is the AI generated discussion of my post, The Sales Process And Contextual Flexibility.  I hesitated posting this.  It is the third attempt to train the tool to give a more accurate answer.  But you can see how it invents and attributes statements to me, which are not in the post.  It has wild hallucinations.  This tells you how far we have to go in the development of these tools.

Why publish, first as an illustration of these challenges.  But interspersed in the lies an hallucinations, are a few gems.  If you listen, listen at 1.5X.  Share your comments with me.

Here is the link to the original post:  https://partnersinexcellenceblog.com/the-sales-process-and-contextual-flexibility/

Doing The Work Or Going Through The Motions?

This is the AI generated discussion of my post, Doing The Work Or Going Through The Motions?  It's actually very interesting, they took a hard left turn in the middle of the discussion talking about the importance of Strategic Mediocrity.  The discussion was very different, but very insightful and powerful.  It went far beyond some of the ideas I had been looking at in the post.  Enjoy!

Here's the link to the original post:  https://partnersinexcellenceblog.com/doing-the-work-or-going-through-the-motions/

A Further Note On Hyper-Efficiency, The Path To Ultimate Efficiency

This is the AI generated discussion of my post, A Further Note On Hyper-Efficiency, The Path To Ultimate Efficiency.  It's short, but they get straight to the point.

Hmmm, maybe they, too, are focused on efficiency   ;-)

Here is the link to the original post:  https://partnersinexcellenceblog.com/a-further-note-on-hyper-efficiency-the-path-to-ultimate-efficiency/