In Blog #26 (Plate Spinning), I mentioned that I was exploring the publication of my collection of global travel stories, Tales from a Small Planet, and that I had been turned down by The University of Texas Press and by Texas Tech Press. I did not mention that my current agent also turned it down. I asked ChatGPT, and it recommended that I explore other academic presses. I upped my effort and have now been turned down by Cambridge University, Yale, Princeton, Chicago, The University of Arizona Press and, for good measure, Springer-Verlag, publishers of my book on Supernova Explosions. ChatGPT recommended some agents and I emailed one with a quick reply of receipt but no further word. I have not had the mental space to pursue other bot suggestions.
On 4/20/26, I read a sample from Tales entitled Bannister to the Westbank Library Writer's Group. This was the shortest of the collection, 500 words, about a time I was nearly caught sliding down the elegant marble banister in the Cambridge Club in London. It got some acclaim from the writers, which was good for my ego, but does not get me any closer to a contract. I did get some suggestions for travel writers I should check out.
Jane, one of the writers, volunteered to proofread and edit the whole story collection in Tales. She emailed, "I made an over-generalization about academics who lacked a sense of humor. You, thank goodness, have a comic awareness." Thus flattered, I could not say "no." I sent her 54,000 words and after a few days she returned detailed edits and eight pages of comments. Absorbing that is now on my agenda.
The Writer's Group selected the theme of "birds" for the Spring edition of the West End Writer's Quarterly, with contributions from the group. I first thought, "I have nothing to say about birds," but after a few days, I realized I could gin up a short contribution. I wrote a piece called Lucky Bird about the road runners that inhabit our neighborhood. They only rarely show, and I always consider them lucky omens.
With fading sales of The Path to Singularity and advice from several quarters, I girded my loins to do battle with the Amazon Ads machine. There is no charge for keywords, so I converted my index into a huge list of them. That was too cumbersome to manage, so I trimmed the list down to a dozen or so while I experimented. I first made very conservative bids and the result, as might be expected, was few free "impressions" a handful of "clicks" for which I had to pay a little, and no book sales. I then tried raising my bid to be within Amazon's recommended window for keywords like Kurzweil, Bostrom, and Singularity. The result reminded me of the dramatic segment, The Sorcerer's Apprentice, from Disney's Fantasia that haunted me for years of my youth. The impressions shot up to over 10,000, the clicks to dozens, and I actually sold a few books. The problem was that while I had set a daily expense limit of $5, I was hitting that every day, spending far more than I was making in sales. I felt like Mickey Mouse in that film, chopping the brooms, only to have them multiply. I tried to pause the campaign, but it kept going. Then I found another pause button, then another. One problem, I figured out in the wake of my panic, was that I had somehow triggered a United Kingdom campaign in addition to the U.S. version. That required a separate pause. I've been working up my courage to tip toe back in. The trick, I guess, is a carefully balanced choice of keywords and bids. I also have to say, I find the Amazon web pages awkward, cumbersome, and inscrutable. We'll see.
This is more about reading than writing, but I had an interesting exchange in April with Ben Thomas, my last postdoc before I retired. Ben is now working at a high-tech company in Great Britain. Ben had engaged in a new personal project related to machine learning. The idea was to take a selection of technical papers on "superluminous supernovae" (SLSN) and make a Retrieval-Augmented Generation or RAG system. A RAG is a technique to enhance AI large language models (LLMs) by allowing them to retrieve and incorporate new information from external data sources (the Web) before generating responses. Ben's notion was to create an LLM expert in SLSN by collecting all the papers and embedding them into a "vector database." Then, when you query the LLM with a SLSN question or thought, it would embed that question-or-thought into the same semantic space and pull out the nearest neighbors that would be semantically similar to the query, thus yielding LLM-assisted science. This project, in turn, required the construction of a "knowledge graph" of papers and their citations.
I'm working with Hungarian colleagues on a paper on the supernova spectral data we have acquired with the Hobby-Eberly Telescope at McDonald Observatory over 15 years. We are specifically exploring mysterious "high-velocity features," HVF, which are frequently observed but unexplained in exploding white dwarfs. In a fuzzy way, I had also been thinking of using AI to help guide our interpretation of HVFs. My notion was to collect every paper on HVF, and query them in some way. What do HVFs have to say about the progenitor evolution and explosion physics? I was thinking of something like a RAG system without even knowing that was a thing. Clearly, I need to learn more about RAG techniques.
Then there is the subject of knowledge graphs. In yet another coincidence, my older son, Diek, gave me a book on knowledge graphs for my April birthday. I have little idea what they are, though I see they are something related to the clustering in machine learning analyses. The book is a collection of technical contributions, and I've been working up my energy to tackle it. Ben's query boosted me in that direction.
I had previously met Chris Barton at an Authors Guild function in Austin. He had suggested that with my book oeuvre, I would be a good candidate for the Texas Institute of Letters. That sounded like a nice idea to me. The hooker is that you cannot apply; you must be nominated. In April, I read that Chris is now president of the Texas Institute of Letters. I emailed him congratulations, noting that I had not spontaneously been nominated for TIL membership, and asking his advice about how I might catalyze such a thing. His response was that one way to get on the radar of other TIL members is to serve as a judge in the TIL annual literary awards in a dozen literary categories, one of which is the award for Most Significant Scholarly Book. I agreed to become a judge and now look forward to receiving 8 or 9 scholarly books to review next October. No self-interested deed goes unpunished.
In the meantime, writing remains an accessible, easy dopamine fix, so I write on.