Tag: youtube

Tripping Up AI with Apple Music and YouTube

It Knows What Tunes I Like. Or Does It?

Many of us who work from home, or are home most of the day, spend a lot of time listening to music. We all have our music apps of choice. For the most part, I use Apple Music and YouTube. I have more albums and homemade playlists on Apple than I can count.

Those of you who have read my posts know I favor ’80s new wave and goth tunes, with classic rock and a few other things mixed in. The mixed-in includes some disco. Yeah, I remember the Disco Sucks pins in the 70s. A lot of it does suck. But not all. Classical, opera, and theater music are also in heavy rotation. Country has never been, and never will be, mixed in.

Other than when I go to sleep, all my music is from the digital platforms. Bedtime is from my CD player. Yeah, they still exist and sound fine. I often let Apple Music or YouTube make a mix. It almost always has mixes I can click on as soon as I log in.

AI creates the setlist. It knows what bands I listen to most often, as well as the songs from those bands. It is not unusual for the lists to be heavy on Bauhaus, Bowie, the Beatles, the

Cure, Erasure, Echo & the Bunnymen, Elton John, Depeche Mode, Love & Rockets, the Ramones, Siouxsie and the Banshees, and a few other bands. You get the picture. I am sure you can pick the rest. Feel free to enter your comments in the chat section.

For the most part, my AI music lists are spot on. But I trip it up. It’s not hard. I often add stuff like ABBA, the Bee Gees, Jackson Browne, “La Bohème,” the Village People, and “Tosca.” Feel free to guess some of the others. You’d be right if you picked Prince, Led Zep, and the Who. All are heavily in the rotation.

Apple often says, “Mitchell’s Mix.” It adds a hyphen and says light, rocking, chill, etc. For the most part, it does a decent job. I guess it is learning my tastes. Occasionally it adds the Grateful Dead. When that happens, I immediately exit the list. I have never listened to them and never will. 

AI is doing a great job with my playlists. However, like all of us, it is still learning. I think it’ll get there. But who knows what I will pop in to confuse it?

AI and Transcript Creation 

For those of you who read my prior AI posts, “The Happiness and Toils and Tribulations of AI” and “What Does AI Really Know About You,”I thank you. Now it’s time to jump into how well it does when it comes to making transcripts when you upload videos to YouTube.

A few days ago, as chairman of the Ten Mile River Scout Museum, I made a few presentations at our annual Alumni Weekend. One was to present a tile in honor of a deceased Scouter who made an impact on hundreds of youths. The other was the introduction of my board of trustees. Both were recorded on someone’s iPhone.

We uploaded the videos to YouTube and transcripts were created. In the first video, “August 8, 2026: TMR Wall of Fame Induction of John D’Arcy,” the only speaker was me. In the second, “August 8, 2026: Presentation of Museum Trustees*,” each of my trustees introduced themselves. We do a lot of work and have fun. We don’t take ourselves too seriously. Most of us have been friends for decades and love to joke around.

I was told by the person who recorded the videos that he placed them on YouTube. I was excited and took a look. YouTube automatically created transcripts. I was aghast when I saw the transcripts. Yes, it got some things right. But it misspelled many things. Before alerting me that it was live, it should have been checked. But it is not the end of the world.

I don’t think YouTube’s AI understands Brooklyn. Everyone in the video is originally from NYC. Many are from Brooklyn. However, as far as the other boroughs, YouTube doesn’t speak that either. 

Several of our camps have Native American names. One of our now-shuttered camps was Nianque. One of our trustees said he started over sixty years ago in that camp. The transcript said, “Camp Miami.” We’ve all been there. Miami and a Scout camp in the middle of Narrowsburg are way different. Personally, I like my camp better. But have had many adventures in Miami. It spelled many first and last names wrong, as well as the titles.

We toyed around with editing the transcript. But it was time-consuming and, for some reason, more difficult on Macs v. Windows-based computers. Why? Who knows? You’d think it’d be easier and quicker on a Mac. Everything is. What did we do? We deleted them and uploaded again without transcripts. 

Am I upset? Not at all. As a matter of fact, I am pleased. This was a great learning exercise. It is fantastic that AI is being put to work to create transcripts. They look nice and are perfectly placed at the bottom of the video. This taught us that you always need to check and verify. I do not doubt that as we move forward, this will get better. My other work with AI tools has proven that. We are on a learning curve, and we are learning, learning a lot.

*- These are the final videos.

Alana’s in Pain

Alana’s looked like she was about to burst. She and Marc were chilling in his camp. Why they were there, no one knew. Even they couldn’t figure out why they weren’t home. The two had no clue where the nearest hospital was. Marc remembered a few hospital runs in camp after they went to Action Park. Someone always got hurt.

They had a mix of Adele, Joy Division, the Cure, U2, and Depeche Mode playing, and even a little Billy Joel, Elton John, the Ramones, and Beatles made the setlist. The fire was burning. The weather was perfect. Marc kissed her. He was very nervous. Their doctor didn’t know of their powers. They had no idea what would happen when the baby was born. They did their best research by reading Star Wars books.

Alana’s water broke. She was in the leanto that Marc slept in during his summer years. She was on the floor on top of a green blanket. The cots were long gone. She started to scream. March began to shake. Wi-Fi was nonexistent in this part of the camp. Due to the connectivity issue, they couldn’t find a video on YouTube showing how to deliver a baby. It was too late for Marc to get her into the car and use Google Maps to guide them to a hospital.

Alana’s face was redder than a tomato. Marc was pretending to be an OB/GYN. She screamed. No one heard her. Or maybe someone did. Suddenly lightning flashed. She cried. Clueless, Marc was holding her ankles. He almost fell. As Marc was falling, a hand pushed him back up. He freaked. He turned around and saw the Woodsman with his herd of buffalo. The Woodsman motioned to Marc, who stepped away. A buffalo waltzed up to Alana. It looked at her. Alana’s eyes froze. She laughed. Suddenly, a boy and a girl popped out. They were perfect.