Up front for those who might be curious, I am re-publishing some very old posts in order to import them into Anchor.fm to automatically convert them to audio. I started using this feature last week when Anchor.fm imported many of my posts, but it stopped for posts that were more than 6 years old. It turns out that some of my favorite posts were the ones I wrote in my first year blogging. To get them into audio (using the automatic feature), I temporarily publish them so Anchor.fm thinks they are new posts so that I can create a new podcast. Once imported, I delete the copied post and then insert the Spotify link into the original post.

I was doing this quite a bit before I was alarmed to see that people were reading the posts as if they were new posts, and even received a like for one of them. Maybe the others also might have earned a like or even a comment if the post weren’t deleted before they finished reading it. I doubt if any one person was tricked into reading multiple such posts, but if there were I can imagine they might be pretty annoyed. Later on, I started to insert at the end of the post a link to the original post, just in case someone did want to respond to a post.

It surprised me to see how many views these old posts received by merely republishing them. The posts have been sitting in my blog for years without a single view since the first day of the post. I figure there is a combination of two factors. One was the rapidity of my posts probably triggered something in the algorithm to think I have a larger site than I do. The other was just that fresh publications do show up in people’s feeds, as is standard in modern times where the interest is on newest postings for the simple fact that they are new. On that point, I feel bad for unintentionally presenting something old as if it were new.

As mentioned above, I was going back to posts 7 years ago to add audio links to them. I recall at the time that the search engines were much more interested in random blogs like mine. At that time, I would see I was getting people seeing my blog being at the front page of their google search results. This happened very often at first. Later, and I am not sure when, this stopped. Not only for my blog, for nearly all blogs. I used to be able enter a search and get blog results, of someone who just happened to be talking about it. Now, my searches do not show blog results no matter how many pages of results I check.

Wasn’t there once a feature where you can pick “blogs” as a search option in addition to “shopping”, “news”, “images”, etc. I don’t see that any more.

There was a time when I could search for one of my blog titles and it would appear in a search result within a few pages, especially when the title was reasonably unique. Other times, I could enter into a search a snippet from my post and find my post in the search results. I am much less successful now.

That said, the statistics page for this blog does show people finding some posts using google or bing. Unfortunately, the search terms are no longer visible. There are only a couple of my posts that do get found this way, and often get found multiple times. In most cases, I doubt that the reader found my content interesting so perhaps the low ranking by the search engine is beneficial for the users. There are a few that clearly do find the post interesting because they will check my about link or some link in the post. Lately about 10% of the readers who find one of my posts from a search engine actually engage with the content. That is a very impressive rate, but I think it is because they only find my content by going perhaps dozens pages of search results because nothing else was interesting. The high engagement rate reflects the few who are really motivated to find a different point of view.

I approached blogging for my own interests. I describe it as diary, a place to just write some thoughts I have, but one where I deliberately make it public. I am not promoting the new posts in any way, but I do hope that someone will find a post that interests them at some time.

Maybe I did start off the first year promoting my posts more. I did get accounts on Twitter, LinkedIn, and other sites where I would post links to my latest posts. These were not very productive so I gave up on them. At that time, again, I was getting more traffic from organic searches.

My hope was that my posts would be what sometimes is called evergreen. These are posts that would be of interest for long time. The posts are not specific to current events or debates. Someone can find interest in posts that were written years ago.

That is true for me as I went back 7 years ago and found the old posts to have aged well. I could have written the same post today and they would seem relevant. That appears to be what happened to those who saw the republished posts in the past couple days: to these readers the posts appeared fresh.

I am a little disappointed that I would have to republish a post to bring attention to it. I had hoped that the search engines would do that task automatically. The original goal, I thought, was that the search engines would become more intelligent in being able to match a unique take of some topic to a reader seeking that kind of novelty. Clearly, the search engines have given up on that goal as evidenced by many search results returning many pages of the same news story that happen to be posted by different large new sites. There is no novelty in the first several pages, all the results are the same story with different bylines.

It does not matter much to me that my blog is no longer showing up in search results like it used to. The experience of the past couple days does at least make me wonder if there might be an audience who would be interested in some of my posts if only it would show up in their feeds like the republished posts did. It is their loss, I guess, if they can’t find something of mine that if found could interest them.

It does matter to me that I can no longer find other blogs for topics where I am interested in views from someone who, like myself, is commenting on some topic informally, just expressing some opinion or just a personal impression about the topic. When I hear some topic that gets a lot of discussion, I quickly bore of hearing what the professional have to say about it. I want to hear what laypeople have to say, especially when they are critically thinking about the topic. It seems to me they were easier to find a decade ago than now.

Just now I thought the analogy of the current progress of the information age being like the cosmological theory of the “big rip” where the dark energy eventually sends all the galaxies so far apart that astronomers would not see any galaxies outside of their home galaxy. This is what appears to be happening in the information age. People are separated so that they can not see each other again.

The point of this post is about why I have been so keen about adding audio to my posts.

First of all, I was very impressed with the AI that converts text to audio. While I can tell that the audio is coming from a machine, the voice, inflection, and variable pacing is very natural sounding, on the human side of the uncanny valley. The advantage of the AI is the consistency in the reading across the entire piece. There is no stuttering or stumbling that beg for jump-cut editing. To be sure, my posts have grammatical errors, and some are quite bad. However, the AI reads through them without stumbling, so that even though it sounds awkward, it is not distracting and I get the basic idea of what he is trying to say.

A second reason is that my posts do not follow a proper organization as an essay or as a news article. I usually just start writing and eventually stumbling upon something interesting at the end of the post. Often, I would write the title after the post is written where the title reflects what happened at the end of the post instead of what started it. When Anchor.fm announced support for WordPress, I immediately felt it was perfect for this kind of blog. Normally, people reading one of my posts would start at the top and then leave after seeing that first paragraph or so is uninteresting. A podcast is different in that it does not take effort to have the voice continue. The voice could even be in the background but eventually the voice would get to the potentially interesting part at the end.

If I were really serious about an audience, I would put in the effort to edit my writing into a well organized essay. Even with the audio feature, it would benefit to organize it as a proper speech. It just turns out that the AI reading of my scrambled notes just happens to be palatable to the ears.

The third reason is that the primary audience is probably myself. It is magical having my posts read back by a different voice and an AI voice that gives the impression of confidence on the topic. I really feel like I’m hearing someone else’s thoughts. With so many posts in spotify, I can have them played back to me as background sounds while I am doing other things, and yet still find some things interesting even though it was my own thought.

The fourth reason is that it fascinates me to hear what I wrote several years in the past. I am not the same person who wrote the posts. I have different circumstances, and different goals now. Hearing what I was thinking many years ago reminds me of where my mind was at the time, and gives me the opportunity to reconsider some old ideas.

Many of my favorite posts were written 7 years ago. It is probably worthwhile to republish many of them, because at the time I was trying to use the blog as a marketing tool. At the very least, I added content to the post to avoid embarrassing myself to people who might consider working with me. For example, I should rewrite many of the oldest posts to rid all the pretense about about deep knowledge about data science. The basic themes about the implications of a data world remain relevant, but I do not need to pretend like I have some kind of reputation in the field. The pretense distracts from the more basic message about how data is changing people and society, and how data will change both in the future.

Another reason for the interest in having AI convert my blog into podcast format is that I’ve been waiting for this. I may not have stated it explicitly, but I was waiting for AI to get to the point where it is in the automatic reader in Anchor.fm. The reader is living up to the site’s name as being like the news anchor on TV reading a script projected on a teleprompter.

Going forward, I expect my future posts will change its character as I begin to think this is a podcast instead of a blog. Maybe, eventually, I’ll even read aloud my own content in my own voice. Or may be not. The AI already is a better reader than I am, and I expect it will be even greater in a few months.

Going back to my analogy. Information technology is separating us individually by demoting our ability to find each other in searches. Now, AI is replacing my voice. It is expressing my ideas, but if there is an audience that audience will have no connection to my person. At least with my written word, there is a personal connection of these letters coming from keystrokes from my fingers. Changing the content to AI podcaster eliminates that connection.