My latest creation of Podcast Notebooklm is deeper and more fascinating than anything I have ever created, and I bet it will shock you too.
I do not understand the theory of strings. In fact, I bet that there are less than 1% of the world which can speak convincingly on the subject, but I am fascinated by the concept and I read a little on it. Not enough to understand you or explain it to you, but enough to have a constant and constant curiosity.
AI, on the other hand, I think that I understand and that I do not regularly use as a tool. When Google has published a recent update of notebooklm which includes, among other things, Mind cards, I thought it was time to bring something on the very external edges of my understanding and this artificial intelligence capacity on a bleeding edge.
So I created a podcast of channel theory.
First of all, a small primer on notebooklm. This is a powerful search tool based on the AI in which you can download sources, and it will transform them into summaries and extrapolated information in the form of text, podcasts and visual guides as mental cards.
For me, the most fascinating song was the podcasts or the “audio glimps”, which produce talkative audio conversations on almost all the subjects you feed them. I call this a podcast because the audio style takes a well -used path from the most popular podcast series. It is conversational, generally between two people, sometimes funny and always accessible.
I wondered, however, if you can stretch the limits of the format with such a deep and honest, confusing subject, that the resulting podcast would be a conversational nonsense.
My experience, however, has proven that although the current version of Notebooklm has its limits, it is much better to understand dense scientific bits than me and probably most of the people you or me know.
Strange science
Once I decided that I wanted to notebooklm to help me with the subject, I went in search of the content of the string theory (there are much more online than you think), quickly stumbling on this 2009 -page research document of 2009 by the researcher of the University of Cambridge, Dr. David Tong.
I scanned the doc and could say that it was rich with details of theory of the strings, and so far, it probably resides somewhere near the rings of Saturn.
Imagine trying to read this document and understand this. Maybe if someone explain That for me, I would understand. Maybe.
I downloaded the PDF, then I fed it in Notebooklm, where I asked for a podcast and a mental card.
It took almost 30 minutes at Notebooklm to create the podcast, and I must admit that I was a little anxious when I opened it. What if this mass of details on one of the most confounding subjects of physics overwhelmed Google AI? Could hosts simply babble inconsistently?
I shouldn’t have worried.
I had already heard these podcast hosts: a pair of vanilla somewhat (a man and a woman) who joke with casualness, while making mind. In this case, they were trying to explain the theory of strings to the uninitiated.
They started by talking about how they would cross the subject, covering bits such as general relativity, quantum mechanics and how, at least in 2009, we had never directly observed these “strings”. Earlier this month, some physicists said they had in fact found the “first proofs of observation supporting the theory of strings”. But I’m getting lost.
The hosts spoke as experts in physics, but, as far as possible, in terms of secular. I quickly found myself wishing that they have a guest. The podcast would have worked better if they were proxies for me, not understanding much at all and had an expert generated by AI in Interviewer.
Turn all together
As the podcast progressed, the hosts dug the details of channel theory, in particular the definition of a “chain”. They described them as tiny objects that vibrate and added: “All things in the universe come from the vibration of tiny strings.”
Things have become more complex from there, and although the tone of the Podcast AI hosts has never changed, I had trouble following. I still cannot tell you what “really means the particle of relativistic points through the special relativity of Einstein”. Although I appreciated the analogy of “imagine a chain moving through space.”
AI hosts used various tips to keep me committed and not completely confused. The male host, like a podcast parrot, would often repeat a little of what the female host had just explained and used decent analogies to try to make it relatable.
Sometimes the female host disturbed in what seemed to have read straight outside the research document, but the male host was still there to bring her back into entertainment mode. He did a lot of talkative summary.
I felt like I was reconnecting when they explained how “the rope turned into theory of everything” and added “bosons and fermions, partners in crime due to supersymmetry”.
It was heavy
After 25 minutes, my head was stuffed to the point of bursting with these still theoretical strings and turning with terms such as “Sommets” and “Holomorphs”.
I hoped a great and glorious summary at the end, but the podcast suddenly ended at almost 31 minutes. He cut as if the hosts lacked flow, ideas or information, and moved from frustration microphones and without signing.
In some ways, I have the impression that it is my fault. After all, I forced these sims to learn all these things and explain it to myself, because I could never do it. Maybe they are fed up.
I also checked the mental cards, which are connection diagrams that can help you map and represent complex subjects such as channel theory. As you can imagine, the Mind Cards for this subject start simple, but become more and more complex when you develop each branch. However, they are a nice companion for studying the podcast.
It should also be noted that I could enrich the podcast and the mental cards with other sources of research. I would simply add them to the Sources panel in Notebooklm and I would replay the “audio presentation”.
A real expert weighs
As far as I learned and as much as I trust the source material, I wondered about the precision of the podcast. AI, even with solid information, can hallucinate, or at least misinterpreted. I tried to contact the newspaper author, Dr. Tong, but I have never heard. So I turned to another expert in physics, Michael Lubell, a physics teacher at the City College of Cuny.
Dr. Lubell agreed to listen to the podcast and give me some comments. A week later, he sent me this brief note: “I just listened to the podcast of channel theory. It is interesting to note, but this requires a reasonable amount of expertise to follow it.”
When I asked questions about obvious mistakes, Lubell wrote: “No evidence, but I have never done research on string theory.” Quite fair, but I’m ready to bet that Lubell understands and knows more about the theory of strings than me.
Maybe the Podcasters Ia now know more about the subject than one of us.
