Ideas from Mark Stout

Read on my Substack  

 I posted this tonight to Reddit , and it is getting some positive traction. We have seen many perspectives about vibe coding in this community.  Professionals are concerned about security, and also seem fearful of encroachment on their profession. I may have a unique perspective. While I have never been a professional programmer, I have been a hobbyist programmer since 1979.  Yes, you read that right, for 47 years. I started on the TRS-80 Model I and then the IBM PC and its descendants. There have been three developments that have thrilled me over those decades: Getting a computer  Going online (CompuServe, Bulletin Boards, the Internet and then the web) And Artificial Intelligence I have programmed mostly in script languages like BASIC, Visual BASIC, PERL, PHP, and Python. And ones you probably never heard of,like Toolbook and Dbase III. And I’ve dabbled in Java, Cobol even, and C. I am not generating SaaS applications with the hopes of getting rich. I do not under...

 I was telling someone tonight about NotebookLM and suggested they check out my video on it. I just realized it is a bit buried it in an old post  about AI, so wanted to surface the video here.

  They say as you get older, two things are the first to go. Memory is one of them. I forget what the other is. Forgive the dad joke, but none of us remember everything, and age only makes it worse. I have a series of tools and strategies to help remember important things. And me, being me, they involve technology. After my medical, err let’s call it a medical “sabbatical”, of 2024, I see a lot of doctors. I’m doing okay, there are just things to keep an eye on. I have taken to asking doctors if I may audio record my appointments. Only one has refused. Once I have the recording, I bring up Google’s  Gemini  AI. You do not even have to write a prompt. Just attach the audio file with the + button, and then hit the Play button triangle to have it run. It will create as summary of the conversation. I pay $20 a month for Gemini Pro, I tested it with the free version, and it would not accept a nine minute long recording for processing. The paid Pro version did. Otter.ai  i...

Ever wonder about how two overlapping things fit in to one another? For example, Railroads in the US in 19th Century, the Steam Engine in the same period, and the US Civil War. I’ve had it in the back of my mind to build a tool to visualize those kind of relationships. I’ve been frustrated though with the seeming lack of a public database of historical dates in a machine readable form. Creating my own would be a daunting task. Then I realized all the dates I’d ever want are already in Wikipedia. I fed Google Gemini AI a URL a Wikipedia page and asked it to give me a list of dates it found in that article. It worked! So I kept dinking around, and had a graphical timeline generated from a URL. Then Gemini suggested instead of giving a URL, I just give it a topic, and it would find the pertinent articles. You can try that tool yourself . <cross posted from  my Substack Newsletter >

About 3 weeks ago I subscribed to a free month of Google's Gemini Pro .  With it, I could make 3 high quality 8 second videos a day with sound.  Three is not many but it served my needs.  And while eight seconds seems low, the next time you watch TV or a film, notice how often and how fast the shot changes.  According to a Wired Magazine article " The average shot length of English language films has declined from about 12 seconds in 1930 to about 2.5 seconds today" Note :  This blog is aimed at two audiences.  One, those not as enamored of tech as I am and the other, my fellow enthusiasts.  For the first folks, my  Substack  newsletter reproduces the relevant posts.  The more techy posts live only here on the blog, as they have for the last 17 years.  This is one of those posts. My project I worked on was a series of about seven unique shots. In each case they were an employee modeling the proper phrasing when speaking to a custom...

Two big things. You can now see my Amazon River notebook using NotebookLM right in your own browser.  With it you may listen to the full podcast, ask questions of it, and more.  It is here . And I fixed an error in the video.  It is here .

 A follow-up to my last post about Google's NotebookLM. Google's developer conference started today.  I watched the keynote and there was a great deal said about AI.   What I thought interesting is they have published a NotebookLM with a lot of data about the conference you can query and summarize in NotebookLM. Sources for the notebook includes a YouTube video of the two hour keynote and a host of articles pertinent to what is happening at IO, the convention. You can access this notebook here .

 Artificial Intelligence is all over the news.  I played with ChatGPT when it first appeared, and I was not impressed.  I also played with Google's NotebookLM.  I was intrigued by it. You could then feed in documents from your Google Drive and then ask questions and it would answer based on what it found in your documents. In the eight months I was away from computers, things changed dramatically.  AI got smarter; and it will continue to evolve. NotebookLM can can now pull in sources from places outside your Google Drive.  For this example, I chose a Wikipedia article and a YouTube article about the Amazon River. It took a few minutes and generated an informative summary of the topic. What was truly amazing was it also generated a podcast about the Amazon, with a man and woman speaking.  Apparently two voices having a conversation are more engaging than one voice lecturing. A link to a video I made showing the demo is at the end of this post. Another g...