Where Love Has Gone

I've always sucked at predicting the future, so mostly I've cultivated a habit of pausing when I feel tempted to do it. In turn, the only positive spin I can put on this fortune-telling deficiency at the moment is that I have plenty of company. There is no shortage of smarty-pants out there, that is, who have inserted themselves into the Artificial Intelligence discussion and what it will mean for us tomorrow. Maybe some of them will get some of it right; maybe not. Some are predicting things that range from magnificently good to altogether bad, and everything in between - and even further to the imminent destruction of mankind. If these predictions have anything in common, at least from what I've seen, is that they are born or nearly so in the mind, and, in turn, most often expressed by scientists, professors and the like, those who engage in mind things often. By the way, this is one thing computers are clearly good at, mimicking the human mind.

In titling this series of recent musings "Where Love Has Gone" I had more in mind - or, rather, more outside of mind and into something else. I'm not sure what to call it, but it's associated with living with a quiet. Koans, the Zen species thereof, illustrate this nicely. Perhaps the most common if tiresome koan of all is, "What is the sound of one hand clapping?," though a close second would be, "Does a tree in the forest make a sound when there is nobody there to hear it?" In any case, the point of generating koans is to challenge logical thinking and lead one to intuitive understanding, something otherwise insoluble with a rational answer. I would put it another way: Koans are not designed to be understood but rather to be experienced by one's (and I'm really searching for a word here that doesn't sound corny) self. And what's utterly essential to experience a koan is to give space for it, as in clearing one's mind.

Many of those seeking some kind of spiritual enlightenment will twist themselves into perpetual contortions looking for this elusive space, even convincing themselves that they know what it is they're looking for, and this is often because they have accidentally experienced an empty mind in their past, saw what it could bring, and unsuccessfully sought to repeat it. Oh, and spiritual teachers can bank millions helping them embark on their doomed journey. Give them things to do. Eat mindfully. Chant. None of these work.

Not surprisingly, many poets seem to get this on some level. Here's Eve Merriam's oft-quoted advice:

"How to Eat a Poem"

Don't be polite. Bite in.

Pick it up with your fingers and lick the juice that

may run down your chin.

It is ready and ripe now, whenever you are.

You do not need a knife or fork or spoon

or plate or napkin or tablecloth.

For there is no core

or stem

or rind

or pit

or seed

or skin

to throw away.

Not exactly something you attack with your intellect.

But enough of this. The question here is, Would AI be any good at helping us to sell books on the basis of giving potential buyers a koan-like experience, insofar as they have the potential of sharing to some extent the experience a buyer would have by reading them? And I'm assuming that this experience would be attractive enough to get them up off their wallets. Surely most of you have seen eBay's wobbly first steps into AI generated seller descriptions. I could cite some egregious examples, but it might be more useful to generalize on the characteristics of what one might typically run into. To this end, Google's AI hits most of these nails on the head:

"The "worst" eBay AI descriptions for books are generally characterized by generic marketing "fluff," a lack of specific condition details, inaccurate information, and repetitive, unhelpful phrases that waste the buyer's time and erode trust.

"Examples and characteristics of poor AI descriptions include:

"Generic, Overhyped "Fluff": Descriptions filled with marketing buzzwords and clichés such as "a wonderful addition to your collection," "a must-have item," or "experience the thrill of...". This "universal blandness" treats every book like a valuable collectible, which is often not the case for low-value or common items.

"Missing or Incorrect Condition Details: The AI does not have access to the physical book and therefore cannot describe actual condition issues like foxing (stains), tears, or missing pages. Sellers often fail to add this crucial human-verified information, leading to listings where the description "says nothing in many words".

"Inaccurate Information: The AI might generate information that is factually wrong about the item's specifics because it relies solely on the title or limited data provided by the seller. This can lead to "not as described" cases and bad feedback for the seller.

"Repetitive Phrases: Buyers often notice the same "word salad" structure and tired phrases across different listings, which makes the descriptions immediately recognizable as AI-generated and low-effort.

"Irrelevant or Nonsensical Comments: In an attempt to pad out the description, the AI might include bizarre or illogical statements, such as describing a denim shirt as "functional & elegant".

"Lack of Essential Information: The AI descriptions frequently omit basic, necessary details that buyers need, such as the book's specific dimensions, edition, or whether a dust jacket is included, forcing buyers to ask the seller or avoid the listing altogether.

Ultimately, these poor AI descriptions are a "red flag" for many buyers because they suggest a lazy seller who either doesn't know anything about the item or has something to hide, making a simple, human-written description of the condition and key facts far more effective."

Of course, the eBay community - so far- has largely condemned AI descriptions. Talk about things that suck. What is important to buyers, including whatever is needed to pinpoint what a book is bibliographically - the bones - but also carefully selected facts that might be of importance to collectors, whether these collectors know them going in or have encountered them for the first time. The latter is by far the most telling, especially if they close in on the stuff of koans.

Can booksellers do this with any book? You can always try. After all, even textbooks have souls. But the goal is to make a few bucks, right? We're stuck on this planet. We have bills to pay, if we haven't been born into money, and believe me, if I hear that cloying advice ever again - "Do what you love doing with your life, and you'll never have to work again" - I might be inspired to choke a what's fast becoming a target of satirists online, Chihuahua. That would gather a crowd, if nothing else. No, work is part of life, and life isn't always painless. And yes, Virginia, there are books you can make money at. Settled? I think so.

But I can't just make a list of them. I've tried. Done it. Many times. Sometimes I think it's cheapened me - or at least attracted a target I wasn't looking to hit. But I can't close the deal for anybody. I can show you what they are, less often where they are, but most importantly, I can't show you how to breathe life into them. That's your job, the "YOU" I referred to in Part II. And I think AI can help keep the wolf from the door.

It's true that only certain books deserve our attention, our intention to build value into them. To repeat what I've come to believe is one of the more important pieces of advice you can give to an aspiring bookseller, it's this:

Tip: The best way to become a seller is to first become a buyer.

This might sound counterintuitive at first glance, on several levels, but here's a corollary to augment the approach: As you shop, look first for reasons not to buy a book. Then, if you don't spot any, listen. Then ask yourself if there is any potential for you to sell it for more than you'll have to pay. Also pay attention to outcomes, especially in auctions. Why do some soar to puzzling heights and some bomb? Sometimes it's luck, timing, intentional or unintentional deception, but in the long run it's presentation.

Presentation can express many positive (or negative) things. A few examples: attention to detail; knowledge; track record and/or seller ratings (aka trustworthiness); photographic skill; a fearlessness to ask for a higher, even outlier price if you believe in what you're selling (this can absolutely attract buyers) and many more, some almost too subtle for words. By the way, part of this fearlessness involves reminding yourself not to be shy about selling your book on the basis of what's cool about it but avoiding, if possible, betraying a fear that you'll turn off buyers with your over-cautiousness. A beautiful method of accomplishing this is to say, "We are not booksellers (implication is that we don't know what the hell we're doing, so we're just trying to protect our asses) so please ask questions before bidding or look at our photos for a description, as if images represent the definitive truth. Or another is fear of unpleasant outcomes, such as backing into the selling process by using a list of seller terms that are designed to filter out certain types of buyers. Common examples of this are: Deadbeat buyers not welcome; serious buyers only; transactions will be cancelled if payment is not received within 48 hours; please contact us first before leaving negative feedback; don't ask us to end an auction early; or my favorite, don't ask us for a partial refund after receiving the book; and on and on. All of these communicate one glaring issue: They are in essence a confession that this seller has had multiple problems in the past, and the odds are higher than normal for problems happening again.

You'll learn all of this and much more when you put your own money at risk. And this is not to say that you should avoid clueless sellers; you should cherish them as long as they're without red flags. These are where most of your opportunities arise. Yes, sometimes you end up with things that don't match the descriptions, but new sellers or those new to bookselling are often bend-over-backwards accommodating out of fear of buyer retribution or being banned from a venue.

Tip: If you're skeptical about the chasm-like distance between outcomes of the same book sold by different sellers and aren't already a customer of Worth Point, register for their 7-day free trial and start by searching for higher profile titles that you know sell for good money, at least sometimes, then marvel at the difference in outcomes.

The differences can be stunning. Next, investigate the descriptions and look for reasons for these discrepancies. You won't always be able to get an answer but often it's right in front of you. Their database goes back years, and descriptions and photos are included.

So, I've been closing in on this AI thing, tinkering with it, and there are a few things I've discovered, assuming you want to lay down your scanner for the moment.

One: Most important of all it can save you massive time. In the past I've spent many hours on some listing descriptions, researching taking up the lion's share of it. It the early days there just wasn't that much you could dig up online. Today there's far more out there, but it still takes time to check various sources, some of which are reluctant to cough much information up, if at all. I've also built a reference library to assist me. This is even slower, but there are many times when it's the only game in town and repays multiple times what a reference book cost me. Vast as the internet is, it isn't everything.

Two: The second thing I've discovered is that AI doesn't always get it right. It often requires some vetting if something doesn't smell right. Happily - and I'm focusing only on Google's AI for now - they supply links to sources of the information they're delivering, so you can go to the source itself and often evaluate it further.

Three: The third thing I've discovered is that you don't have to struggle to frame your questions just so to get a good answer. It seems to have the ability to discern what you're asking, however awkwardly, and go from there. And, yes, there are times when it doesn't quite get what you're looking for but will offer some alternative questions that will deliver it.

Four: The fourth thing I've discovered is that you can do anything you want to with it. If you do nothing more than copy and paste something into your descriptions, there are no copyright restrictions whatsoever to prevent you from doing this. Of course, more of than not I'll do some editing, light or heavier, when it seems to drop into one of the "worst" categories listed above. Your bookselling "voice" is important, and you can often recover it with minor changes in diction, etc.

For now, I sense this article is getting long, so I'll wrap it up. In my 23 years of providing resources for booksellers, I've learned that what's almost always useful, if not essential, in driving a point home is providing concrete examples. Case studies, in this case, of how AI can build value into specific books. Look for these in Part IV.