Update Announcements by Craig Stark 5 February 2026

By Craig Stark

In the last month or so I have investigated AI and bookselling integration with a attentive focus on what this could mean for us. I'm sure many of you now know how flawed the specifically eBay attempt to use AI in building descriptions of your books (or whatever you sell) has been a colossal failure. If you haven't used it yourself, surely you've seen it used. And I bet you've scoffed. Scoffing is what educated people do - and this especially applies to booksellers, who tend to know a thing or two, when they express damning displeasure. My take is that, if anything, eBay's use of AI degrades your chances of selling anything near its potential outcome, that is, if compared to the result of a professionally written and illustrated description on the same item. I recommend counseling for whichever focus group approved its launch. And no, anybody's light editing of their AI descriptions doesn't change my opinion.

Anyway, I've recently looked deeper into AI, e.g., Google AI, Chat GPT and a few other AI generators, and I've come away from this more alarmed than ever. We're told that these tools will be self-correcting over time, sort of like Wikipedia, I guess, which, to be fair, has upped its game in recent years and improved user trust significantly from the time it was founded some 25 years ago. Still, Wikipedia is far from done, though most of their work looks to be needed most at what they refer to as stubs (incomplete articles) more so than correcting inaccuracies of more comprehensive entries. My hope is that the pace of life will continue to accelerate along with their growth as well as AI's.

In today's feature article, Part IV of my series "Where Love Has Gone," is a case study of a book that looks in part at the trouble you can get into using AI in its present infancy. And stay tuned. The plan is to do two more case studies in the coming weeks.

One final note … a copy of BookThink's detailed history of publisher Grosset & Dunlap is now available as a print booklet. Its history is important for two reasons. One, there are millions of G&D's out there both in commerce and the wild. And, two, it has been denigrated as a reprint publisher for as long as I can remember as not worth-bothering-with and yet, not only did it publish a large number of what were in fact first editions but also produced those first editions and reprints that can easily reach three numbers, sometimes four numbers - if.

The "if" is why I wrote this thing. Along with G&D's being often dismissed as worthless, there is this small matter of identifying publishing dates as opposed to copyright dates. If you read BookThink with any regularity, you already know that publication dates and copyright dates are two different animals. If sometimes they match, fine, but it brings a world of confusion into accurately identifying instances when they either don't match or the publication date is entirely absent from their title. The latter, with respect to G&D's, is sadly more often the case than not: a copyright date only. And if anything is more tempting to a newly hatched bookseller (or even an ignorant one), it's to assume that the copyright date is an issue point that guarantees it's a first edition, even if the book in hand was published 50 or more years later.

But it's not hopeless. What this BookThink Guide will show you is that publication dates, in the majority of copyright-date-only cases, can be established with certainty.

Headings include: "Essential History of the Honorable Pirates," "An Historical Chronology," "How to Date Grosset & Dunlap Publications," using mostly freely available online available references, and "Specialty Resources," a list of reference material you can add to your private reference library, depending on your area of focus. In all, it's 42 pages of tools to add to your sourcing arsenal, along with many color illustrations. Interested? Buy it It's available in print version only here.