Wednesday, July 16, 2025

Reviews in the age of AI

 

This year, over a twelve month period I am (God-willing) launching a dozen books (most of them completing work/coauthored with my late father, B. Ward Powers — nine of those appearing in the two Rising collections). This means I am knee deep in marketing, with a particular focus on seeking reviews and putting books in for awards.

However, many of the reviews I am getting feel deficient — and there seem to be two main reasons for this.

1. If you are reading on Kindle Unlimited, it only counts as a Verified Purchase Review if you read it steadily (at what Amazon regards as an appropriate reading speed) and complete it on your phone or ereader (which involves a popup keyboard and some sort of AI assistance). And all Amazon requires is twenty words - and many reviewers, particularly older people, hate typing on phones/pads.

2. We now have readily available LLM-based AI-seeming chatbots, as well as predictive text, speech transcription, and readily available AI-help directly there in your phone if not actually a button on your popup keyboard. It is naturally very tempting to use this AI help - and even if you are just using word suggestion, auto-correction or speech-transcription, that is AI help too. So it is hard to avoid AI.

As an AI researcher, and pioneer in the development of natural/neural language learning models (NLL/LLMs) and embeddings (in the 1970s to 1990s), I have hundreds of publications on AI, and have blogged about these LLM developments recently:

https://martiward.blogspot.com/2025/03/putting-science-back-into-sf-ai.html

https://martiward.blogspot.com/2024/05/ai-are-we-there-yet.html

I am thus writing not an AI naysayer but rather an AI developer who is currently researching how best to apply and improve AI, and in particular how to use AI/LLMs in the context of health support, social media, and as an author. 

And in this blog I will turn my attention to their use in reviews — and I for one am drawing the line at using it to write reviews, or even to rewrite reviews.


What's in a Review

I've received approaching 100 reviews this month, (editorial reviews and reader reviews), particularly focussed around the two BWP+MW Rising collections, as well as my Time for PsyQ (and its fully-cast AI-performed audiobook, plus Apple and Amazon AI-narrated audiobooks of the entire backlist).

I read every review and I've actually been through all the Time for PsyQ reviews (over the last two years), and marked them up according to the following idealized structure of an effective review:

Author@ - @ signifies that evidence of personal engagement and impact is missing (AI-generated?)


Abstract - pithy statement covering the book's purpose/nature & identifying the ideal target reader.
Objective spoilerfree overview of the book & how it addresses/develops the theme and characters
Subjective view of how it appealed to me personally as the reviewer (and where I stand vs target)
Identification of any negatives, suggestions and critique (put as positively and helpfully as possible)
Verdict - drawing all the above into a succinct conclusion/judgement & targeted recommendation.

Before I explain these, I should first note that there are at least three quite different kinds of review:

1. Developmental review - in which I lump input from alpha/beta readers, copy editors and proofreaders;

2. Journalistic review - which includes anyone who reviewed a free copy/ARC, including bloggers;

3. Purchaser review - which means anyone that paid for the book and provided an unsolicited review.


Review Mismatch

Some club type reviewers don't quite know where they sit, and indeed some clubs expect their reviewers to enumerate the errors (or at least the first ten) and deduct stars according to the number of errors found (in some cases these first ten will be anonymously provided to the author). That is they are really in category 2 as they have been given an Advanced Reader Copy (ARC) but are acting as if they are in category 1 (e.g. playing the role of copyeditor and proofreader despite not being qualified as such), and then as if they were a purchaser they put this up on Amazon and/or Goodreads as if they were in category 3.

You really don't want these types of review mixed up, and indeed the final stages of developmental review (technically a form of editorial review since actually helping you edit your manuscript) are usually done in parallel with the journalistic review (also a form of editorial review but actually they are book review editors for their platform not for you as author). Print magazines and newspapers generally want a lead time of at least three to six months - and if it is a quarterly professional book magazine it is likely to be well over three months and nearer the six months that is needed).

That is both of these processes (1 and 2) start at least six months before the book is final (and possibly even before the cover is done).

A variant on this is submissions for awards, where often the deadline closes off before the end of the year covered, and so if you have a late publication target (e.g. Nov/Dec for calendar year awards as most are) then again you will be providing a manuscript without the final formatting, proofreading and cover - in this case it could be provided in double-spaced standard manuscript format to emphasize it is still a manuscript, and I also include riders about its status in such editorial proofs and review ARCs. It is a mistake to send them your early proof copies as if they were published works.

So this is what should happen

1. Developmental review/beta/proofreading send their review/edits to you only (and are normally paid);

2. Journal reviewers/bloggers should ignore superficial errors in their ARCs (or send privately to you);

3. Purchaser reviews should concentrate on the impact on them as a genre reader (but all is fair game).

The 5-part structure I have outlined above applies to primarily to journalistic and purchaser reviews, but a developmental review may follow it as well as a kind of SWOT analysis to show Strengths and Weakness, Opportunities for improvement and market/marketing Threats as context for providing detailed suggestions as to building on the strengths and opportunities and overcoming the weaknesses and threats.

In particular, note that it is not appropriate for editorial reviews of type 1 and 2 to publish comments about technical errors, and normally positive comments from such reviews are selected by the author/publisher for use on their website, backcover or publicity/press releases (or put up by the author/publisher in a special section on Amazon using Author Central).

The whole purpose of review type 2 is to have pithy endorsements FOR your cover and marketing material AT launch. So you can't possibly ever give them the final published book to review given the aim is prepublished comment to use on the book cover and as part of the launch publicity (publication = publicity not mere printing).

Awards can vary: some do take manuscripts (and indeed may require prepublication access) while others judge published books (and generally have closing dates around the START of the year following publication). These are very different beasts, and indeed there is a third, the above two approaches to professional awards being more like category 2, while a third class of reader awards is more like category 3 (and will often close near the END of the following year, or have a nomination period associated with a midyear event where members/participants will vote - think Hugos/WorldCon).

So the moral of the story, is to keep the three kinds of review (and awards) separate — whether you are author, professional editor, professional reviewer or purchaser/reader.

As noted above, the prepublication versions of the manuscript should generally be in a standard double-spaced format (see https://www.shunn.net/format/story/). The wide-margin double-spaced manuscript traditionally allowed editors/proofreaders to write corrections between the lines with comments/codes in the margin. The wide spacing also has the effect of slowing readers down so that they can focus on technical issues without getting distracted by the story (moving a ruler down the page can help here too).

In the past, such manuscripts or galley proofs (https://en.wikipedia.org/wiki/Galley_proof) could also be used for advanced publicity/review purposes. These days, Print-on-Demand (PoD) allows Advance Reader Copies (ARCs) to be presented in standard book format — but it is important that reviewers know they are not the actual published work, and for this reason Amazon KDP adds a "Not for Resale" banner across the front cover (which itself may be just a mockup rather than the final/commissioned/accepted cover). It is also good to mark this clearly throughout the front matter, or even with a running header.


Review Structure

I am henceforth primarily focused on published purchaser/reader reviews (so exclude the development, editorial and proofreading phase as totally separate animals, avoiding those commercial 'book clubs' that mix the different types of review). These are inherently similar, although the ARC/proof reviewer must also avoid comments that can only properly be made if/when they have the final published book.


Abstract
— pithy statement covering the book's purpose/nature & identifying the ideal target reader.

Your first duty as a reviewer is to the reader of the purported genre of the book.

If you have a blog or a magazine or a newspaper column, this will relate to any specialized niches of your publication too: you will not only give the title of the book but the genres that are consistent both with the publishers marketing and your own views and target readership.

Many readers will take in the cover and title and author of the book in about a second, and then spend a second on the title of your review (which is usually a pithy endorsement) and the first line (which will explain the nature of the book and what kind of readers will be interested in it).

This is in many ways like the concluding verdict - and I have avoided use of the word 'conclusion' as the keyword because it has come to simply mean 'ending' in a very weak sense of being the last paragraph or two. But 'conclusion' is actually what you 'conclude' as you draw threads together at the end of your assessment, and balance the pros and cons of what you've been discussing. And even 'end' also has overtones of purpose and completeness, not merely running out of space or time.

Academics, students and speed readers are taught to pay special attention to the opening lines and concluding lines because these bracketing summaries should tell them if/why they should read it and summarize why. Sometimes only the abstract is published or only the first few lines are visible and you actually have to click (and often pay) for more. In this case, it should also summarize the conclusion. If not, my preference is not to give away the ending at the beginning...

The abstract/endorsement is pithy and will include some nice descriptors - and might just end up on the cover of the book or in the publisher's extended blurb. Professional ARC reviewers often deliberately provide such short pithy separate endorsements. Generally, they should be a single line (<64 characters), although for social media often 100, 120, 140, 160 or 175 characters is the limit - including the attribution to the reviewer and source publication. Normally it will be written in the third person (readers/authors) or second person (you)  — rarely would it be in first person (I).

Have a look at the opening sentences of a few reviews, and note the different character we get with the very few first person and second person intros. I've selected some below for a book that is generally viewed as middle grade, but I regard as being for young adults of all ages (or ages 9 to 99). Note how many are careful to identify the target readership (tween, teen, middle-grade or young adult), and in some cases indicate that it is not just for those younger readers. Others focus on its educational role.

Examples (from Amazon reviews of Time for PsyQ):

A very interesting read, it mixes science fiction with psychic development in an original and entertaining way. 

“Time for PsyQ”, the first entry in the Quantum Talents line of books by Marti Ward, is a fable couched in a young reader's science fiction skin. 

Time for PsyQ is a fun, clever read about Airlie, an 11-year-old who discovers she can read minds with help from her cat. 

Time for PsyQ is an engaging and imaginative middle-grade sci-fi adventure that skillfully blends mystery, magic, and a touch of quantum science. 

I think this is a great book to introduce kids to the next level of reading, going from kid's books to chapter books. 

Time for PsyQ (Quantum Talents Series Book 1) by Marti Ward is an engaging adventure that blends science fiction with a touch of mystery. This book offers a fun mix of supernatural elements, friendship, and personal growth that makes it easy to get lost in its pages.

If you are looking for an exciting novel for your adolescent with an intriguing plot this is the one that you should pick. 

Marti Ward’s Quantum Talents series kicks off with Time for PsyQ, a captivating adventure designed for young adult readers. 

This novel is a good example of how the imagination works in everyone's minds. In “Time for PsyQ” Marti Ward invites readers to enjoy an exciting adventure of teenagers.

This title made me feel like a child all over again, believing into the impossible. Anyone can enjoy this story about a shy girl who discovers she has a unique set of skills, which she can use to help others. 

Very surprising read, a fun one! I really liked how things start with a little mystery around the school and then Airlie, a very smart and curious girl, starts solving crimes and matches in other schools with kids like her. Although my niece just turned 8 years old, she absolutely loved it! 

Reading this book felt like becoming a child again and realizing that the impossible might actually be possible. 

Even if this book targeted children as readers, as an adult I still found myself enjoying it. 

TIME FOR PSYQ, the debut novel in author Marti Ward’s new QUANTUM TALENTS series, is a fast-paced and exciting adventure story for older children and adults and would be an excellent option for reading aloud at school or home. 


Objective
- spoilerfree overview of the book & how it addresses/develops the theme and characters

So between the bracketing abstract and conclusion, as in any structured piece of writing, we need the content, and we should start objectively with the content as the author intended to convey it, without any spoilers (that is important details not in the first 10% or so of the book - which is often available as a preview anyway). On the other hand, it should not simply regurgitate the author's blurb and publicity, and especially not other people's reviews.

This is what the AIs tend to do, and the lack of anything that isn't provided publically by the author/publisher tends to suggest you haven't actually read the book (or that you are an AI).

So although it is objective, it has been through your brain, and taken into account who you are and what the genre is and what audience/demographic you represent, and casts it appropriately. An editorial or journalistic review may give a bit more of a synopsis of the whole book (while avoiding giving away the ending or other surprises/secrets), but review that shows on the purchase page is a lot more immediate and must be more careful not to steal the author's thunder.

So what will distinguish this objective overview from a recap of the publisher's blurb? Largely it will be in the way you describe things, your own style — even your sense of humour — can show through, the adjectives you use, the precise phraseology... Sometimes you might give a perspective that the author wouldn't have though of — although hopefully will still be able to recognize and see as useful.

It may also include information about the author and their goals and perspective, or where it fits in relation to their corpus of writings. And don't forget about including interesting unique keywords and comparisons that will help the right kind of readers find it - publishers and bookshops really like comparisons.

Here again are examples for you to consider, chosen not because you should emulate them as particularly good, but for diversity — so you can think about what your unique way of presenting a plot might be. Being yourself and bringing new variety to the reviews is more important than any particular idea of 'good' (and indeed Amazon is likely to start filtering reviews that look too standard and generic — reviews are now taking longer as it increases the rigour of its filtering).

Examples (from Amazon reviews of Time for PsyQ):

Airlie Sanderson's eleventh birthday introduces her to a Brain-Computer Interface kit and a mysterious kitten named Sooty. As Airlie explores her newfound psychic talents, she becomes embroiled in a local mystery. 

Focused on the adventures of the main character Airlie, her cat, Sooty, plus her friends and their pets, this story begins with a pet theft. As she attempts to help solve the crime, Airlie discovers that she has a mysterious ability to review events in extreme detail in her mind’s eye. Her talents attract the local police, who recruit her into a specialised camp for other kids and their pets that show special extrasensory abilities.

Airlie's discovery of her psi abilities leads her on a thrilling journey, solving mysteries and navigating a world of intrigue. The author created a compelling narrative, with originality and enchanting twists.

This book followed the adventure s of a teenager as she used her ‘perspectives’ or ‘visions’ to help solve problems, meanwhile getting noticed by a Police Inspector and offered numerous opportunities because of them. It was quite the interesting story. 


Subjective
- view of how it appealed to me personally as the reviewer (and where I stand vs target)

Subjective simply means what you thought of it, and how it affected you. It is also an opportunity for you to agree or disagree with the author or other reviews (although we deal with overt criticism of the work next). I tend to look at the middle of the range reviews before buying a book, and sometimes again before writing the review (especially if I want to disagree with a reviewer). 

Sometimes reviews are off track (I've reported several to Amazon that are clearly about a different book, or get things objectively wrong - but nothing has ever happened). So think of reviews on Amazon as a social media dialogue, not just an objective summary. Think of them as being in context not just of the reviews of this book, but of the author's other works.

Sometime we don't have time to write a review, or think we couldn't say it better than another reviewer already has. In that case you could say, like XXX I think YYY — or you could simply upvote them (which should mean their reviews rank higher and display earlier in the "top reviews".

Examples (from Editorial and Verified Purchaser reviews)

My middle-grade reader never felt overwhelmed, and any questions that arose gave us an opportunity to check it out together. The standout for us was the use of a Braille Neue BCI keyboard, and the teamwork in the coordination between characters. Airlie’s technological innovations and the subtle communication methods employed are a testament to Ward's careful attention to detail. 

The appendix, "Science for PsyQ," is a valuable addition, offering readers a deeper understanding of the scientific principles behind the story. It bridges the gap between fiction and reality, making the book appealing to science enthusiasts and curious minds. While the book is packed with action and suspense, it also raises important questions about ethics, privacy, and the societal impact of emerging technologies. 

I liked that it's not all about the action, but about personal discovery and the possibilities the mind offers when connected to something greater. At times it feels like an intimate talk about human potential, wrapped in a futuristic story. Ideal if you're drawn to mind powers, ethical dilemmas and quantum mysteries, but also looking for a hopeful message.

I love exploring young adults books sometimes. They can be a great escape but also not overwhelm me with plot lines and character development.

I liked how it mixed science ideas with mystery in a way that felt believable but still exciting. Airlie’s curiosity and determination stood out, and there were enough twists to keep me turning the pages. 

What stands out most is the strength of the character dynamics. Airlie and her friends form a tight-knit crew, their relationships brimming with warmth and authenticity that make you root for them from the start. The story’s pacing keeps you hooked, with Airlie’s investigations leading to a thrilling race to save a friend at a summer camp for kids with similar PsyQ powers. The addition of a mysterious scholarship offer to a school with “opportunity classes” in quantum psychology is a brilliant touch, sparking curiosity about the broader world and Airlie’s place in it. The novel’s Enid Blyton-esque charm, as noted by The Wishing Shelf Awards, shines through in its cozy yet suspenseful mystery-solving moments.

I first came across Marti Ward’s books when I read Moraturi Lost, which forms part of his Paradisi Chronicles series, a series I also recommend. Time for PsyQ is something quite different and still an interesting read. As the title of the book suggests, the author’s background in behavioural and cognitive/ psychological sciences shows a little even in this book, even if aimed this time at a younger audience.


Critique
— 
Identification of any negatives (put as positive sugggestions as helpfully as possible)

Yes! I'm going to include the full set of negatives from Amazon. But my main point here is to be fair and gentle — particularly towards the indie author and their first book. Try to be helpful rather than critical, and definitely don't try to score "my grammar is better than yours" points. On the other hand when a famous traditionally published author seems to have gone off track, or just be churning out tripe around the same old trope, don't hesitate to say so.  I am likely to identify the same issues in a very different ways to these types of authors. The best-selling author and publisher can take the sharp judgement, and their readers deserve to know that this book is not up to scratch.

But remember, the average indie author sells less than 100 copies — mostly to friends and family and by word of mouth. They might make $100 if they don't pay for any advertizing or covers, or developmental editing, or line editing or copy editing or proof reading — all of which cost $1000s at standard commercial rates (this is all in house for the traditional publisher). We want to encourage them to write more not less, and support them in that with positive critique about how they can improve (and with eBooks and Print-on-Demand they can actually fix the book and republish quickly - in which case a sharp negative review will look forever out of place).

Also remember that there is a difference between gently drawing to the writer's attention a frequent typo or misused word, and presenting your own opinions about plot, characters or structure. Often this could be expressed as suggestions or desires ("although I would have liked to see more development of the other characters"). In fact, comments like this can be quite encouraging (and encourage me to write that sequel or alternative PoV novelette - in the case of Time for PsyQ, there is a story in Fantastic Schools, from the PoV of the Police Inspector, and there is a sequel in progress, School for PsyQ, that parallels/expands it).

Also, do check that that word or punctuation or usage you are correcting isn't actually already correct, or idiomatic/traditional in some other dialect of English. This is particularly important for grammatical issues or spelling variations in dialogue, which of itself can be intended to tell us something about the speaker.

Also be careful about Political Correctness complaints. And remember the author has a voice too in their blog etc. and can potentially make you look stupid. I responded once (indirectly and without direct reference to the review/reviewer/masthead) to a comment in an editorial review complaining about me describing Kate as blind: https://martiward.blogspot.com/2023/12/challenged.html

Note that if you aren't giving four or five stars, it is important to give an indication as to why (although this can be subtle: 'damning with faint praise'). Or some reviewers preface their reviews by explaining that for them it takes an exceptional book to get five stars (think distinction or high distinction: DN/HD).

Examples (from Amazon reviews of Time for PsyQ):

After a slightly sluggish start, this story hits its stride somewhere in the second act.

There are a few predictable elements, but some clever twists help to keep the story balanced.

My only critique is that the summer camp and the other PsyQ kids could have been explored more deeply — I was craving extra scenes showcasing their powers and group dynamics. 

What I'm struggling with is why this book was split into three parts. I understand that each part represents a portion of Airie's story, but it didn't flow together. We see her have friends, lose friends, make friends, etc., but nothing was resolved or understood as to why friends come and go here. I think it would've been better if each part was its own book and just expanded further. Many of the characters who were introduced later in the book deserved better development.


Verdict
— drawing all the above into a succinct conclusion/judgement & targeted recommendation.

Now we return to the objective voice, draw our SWOT threads together and sum up into a recommendation focussed on who will like this book - paying attention to any special features or additional genres/audiences that it reaches (and in the case of children's books, their parents and teachers).

Examples (from Editorial and Verified Purchaser reviews)

Time for PsyQ is a compelling read for fans of science fiction and young adult mysteries. Its blend of scientific exploration, emotional depth, and thrilling adventures makes it a standout in its genre.

Overall, this is a brilliant book with positive messaging, a bit of mystery, and a lot of fun. Very highly recommended.

I think that is the one that sixth through eighth graders and even adults will actually enjoy. I bet that this is just a good book series to come and more.

This is a true sci-fi novel that is worth exploring for your average reader while managing to keep your interest without being too long overall.

Blending elements of mystery, science fiction, and coming-of-age, this book is both fun and thought-provoking. With unique subjects like Quantum Psychology and “PsyQ” powers, the story opens a door to a world where science and magic blur. Bonus links to real-world science add depth for curious readers who want to dig deeper.

PsyQ delivers a zippy pace with an expertly charming voice, enough emotional conflict to give the characters weight, and mystery and adventure abound as well. Resonant of Harry Potter and Scooby-Doo, but grounded in real science education that will engage even the oldest of readers, the narrative strikes just the right blend of adventure, mystery, humor, and moments of soberness exploring friendship and identity.

An exhilarating, educational and heart-felt adventure that will have readers clamoring for more in this universe!

Ward does a good job of exploring the themes of friendship and courage. The book is mainly targeted towards young adults, but some older readers may see something of their younger selves in Airlie.

Beyond the fun, the book weaves in deeper themes that resonate on a universal level. Airlie’s journey of embracing her unique gifts mirrors the broader struggle of finding identity and purpose in a rapidly changing world. These subtexts elevate the story, making it thought-provoking for both its young target audience and older readers. The inclusion of online appendices tying the story’s science to real-world concepts is a clever nod to curious minds, though I didn’t dive into them myself.

If you like tales about young heroes finding their place and their powers, this one will feel like a cozy escape into a world full of wonder.

I highly recommend this book and I would love to read a second part.

This book makes a great gift for the summer break!

The story is a fun mix of adventure, mystery, and science. I would recommend it to any reader who loves a good story and wants to dream with their eyes wide open.

Perfect for sparking imagination and a love for reading.

A great read for fans of imaginative sci-fi adventures!

Perfect for readers who enjoy exploring the intersections of science, psychology, and possibility.

This is a great book for preteen-aged kids.

If you have a young reader, this would be a good book for them. I would recommend this book to any young reader.


AI Reviews and Shortcuts

I started this by mentioning that I'd seen a lot of reviews that seemed like AI reviews because they didn't go beyond recapping what was on the Amazon and Goodreads pages, and often plagiarized other reviews.

If you are using a phone you could put in some keywords with your thoughts and the book title and get an AI chatbot to do something a bit better than this. That would at least have your ideas and reactions in it.

Or you could write your review and ask the LLM to correct/improve it. This would even better match your ideas and views, just correct grammar, spelling and may improve it stylistically. As long as you don't allow it to expand it. These LLM chatbots do tend to be very verbose... So set a target word count.

The App shortcut

I think one of the biggest reasons people are taking the AI shortcut is how difficult it is to write reviews on a phone - and Amazon won't mark it as a Verified Purchase if you got it from Kindle Unlimited and didn't read it slowly and steadily in the app and do your review in the app.

Unfortunately for me, it has also meant that I am frustrated as I try to compose the review, and tend to to be a little short in more than one sense. So I have to work against that - and indeed will try to come back and edit on a computer in a more user-friendly interface, without the pressure of completing it in the app.

Indeed, there is a fairly direct work-around for this: you can install the Kindle App on your computer, and upload the KU book into both. So when I have finished a KU borrow, I swap to my computer to write the review comfortably on a full-sized keyboard and screen. Of course, I can still edit the completed approved review with the benefit of hindsight and having slept on it.

You should also be able to write the review in a notepad or word processor that highlights errors and offers corrections of spelling and grammar, and then past across into the KU app. Although sometimes, "rich" text from an office suite gets converted to images, and needs to be pasted via a plain text editor.

Hope this helps, and avoids you getting tagged for non-contentive AI-like reviews — as Amazon is doing. Just this week, a reviewer of some of my books got kicked out of Amazon (although the reviews of my books seemed fine to me, short but structured along the lines I advise in this post). 

Reviews with a cancelled account can still show up as 'Amazon Customer' — so don't call yourself that as it suggests you have done something naughty!


Good Reading/Reviewing,

Marti
Ward



Note that the Full-cast audiobook is available via the above link, but may also link the Amazon/Audible and Apple audiobooks, which are their own AI-narrated versions not the Alistair G (for Google) version with the full cast of supporting AI voices. The main advantage of the Amazon version is that it syncs with your Kindle if you want to start it at home and take it with you on the road...

Thursday, March 20, 2025

Putting the Science back into SF: AI

I'm leading a full day workshop on SF on Saturday for WritersSA (State Library, Adelaide, SA) that includes 5 session with both "lecture" and "practical" components.

Science, Magic & Speculative Fiction

https://writerssa.org.au/event-registration-civicrm/?id=358

SESSION 1: AIs AS TOOLS AND CHARACTERS (45 MINS)

Discussion on effective research methods, using AI tools ethically, and avoiding common cliches and AI tropes in science fiction writing.

SESSION 2: ENGINEERING YOUR WAY OUT OF PROBLEMS (45 MINS)

Exploration of theme and plot development, world and character building, and solving problems using scientific principles.

SESSION 3: WRITING AN AUTHENTIC SCIENTIST (45 MINS)

Overview of character motivations, goals, and conflicts specific to scientists.

SESSION 4: WORLDBUILDING AND CHARACTER ARCS (45 MINS)

Discussion on how to develop compelling worlds and integrate them into character development.

SESSION 5: SCIENCE AS MAGIC + MAGIC AS SCIENCE (45 MINS)

Discussion on the balance between mystique and plausibility in speculative fiction; Arthur C Clarke’s idea that any sufficiently advanced science/technology is indistinguishable from magic; Marti Ward’s Appearance of Magic™ universe.

LLMs as AIs

One of the trickiest aspects of this (Session 1) is how to deal with AIs - and how today's so-called AIs deal with the various manifestations of AI in science fiction literature and film.

In particular, currently LLM-based AI chatbots have vague purposes, but Copilot has recently been incorporated into Office365 (both the products and as a price increment in your subscription).

Basically in that context it is designed to help you get started on your document or restructure it. Originally however, Copilot had a reputation for being able to write code for you, and tabulate data. It still can, although as risk management constrains it what is does is increasingly restricted/poor.

One of the issues, as an author, is that it can plagiarize work. The more well-known and cited/quoted the work, the more copies it has seen, the more likely this is - and of course the more likely a user explicitly tries to elicit quotes from a work. Again with increasing risk management, this direct copying and copyright violation is largely being eliminated in the big name products. 

The problem with Large Language Models that are too large is that they go beyond being language models to being general repositories of human knowledge and culture, and that includes the stories that have become part of our culture, and it also includes all the social commentary and misinformation available on the internet today. Again, a risk that needs management but is not adequately dealt with. 

I try to direct it (prompt engineer) to provide only information that it can find in refereed sources, and provide those citations in a formal way. But they still point to lots of websites.

Several of my own books are known to them, and it can give the plot and discuss characters sensibly. But I've not found any tendency to regurgitate the texts.

LLMs vs AIs in fiction

The flip side of this new development in AI is how primitive the AIs in our classic movies and shows now appear. Interestingly, HAL holds up well - illustrating some of the risks that our current crop of LLM-based AIs exhibit. Although my Casindra Lost AI, AL, is quite deprecating in his discussion of HAL - regretting the similarity of names and the bad rap/rep that gives him.

One area that HAL and other AIs in fiction get right, and current LLM AIs still get wrong, is 'grounding' - that is being the brain of a robot or vehicle that allows it to fully understand the world in terms of the way it can interact with it - and to learn when new parts of the world expose itself to its sensors.

Really LLM AI Chatpots are not the pinnacle of AI, our autonomous vehicles are. The GAN-enabled AIs dealing with sound and/or images have an edge too, although the understanding isn't there in the same sense that it is for a self-driven car or plane. On the other hand those systems are limited to what is needed for their driving tasks. Although gradually control of sounds systems, integration with smart phones and entertainment systems and climate control, etc. is expanding on this. And LLM AI is starting to join the party.

Did you know?

The prove-you’re-human Turing Tests on websites are actually used to train AIs, and they think they know around half: as the AIs get better, the tests get longer and harder.

This is feeeding into the above systems. This grounded type of system is where AIs will reach maturity and understanding and sentience - in the sense of sensing and understanding and reacting to the world.

Lost Missions

My Paradisi Lost Missions aims to put realistic research in AI into illustrating what we can expect of AI in the stories. While LLMs have hit mainstream since the series appeared, it is based on my 50 years of experience pioneering Learned Language Models - with what constitutes 'Large' growing exponentially year by year, growing by a factor of ten or more each decade — now boosted by the resources of some of the largest companies in the world.

Casindra Lost's AL is introduced as an already fully functioning AI that develops a personality over the course of the book.

The Moraturi arc is a kind of sequel, but focuses on another mission where the main quantum computer and AI are out of action, and other lesser AIs must taken on more responsibilities than they were originally designed for.

    Amazon book pages for Casindra Lost

    (tiny.cc/AmazonCL – Kindle e-book; paperback: ISBN: 9781696380911)


    AI-generated Audiobooks and Translations

    There is no way I could ever recoup the cost of paying someone to do an audiobook or a translation. And narrating myself takes time away from writing (and since I make a few fluffs it takes about twice as long as the final book). I did try to do accents and characterizations for Time for PsyQ (and the first couple of chapters aired in this form on Antipodean SF), and also did some pitch changing to improve the child and female voices (finding that that didn't better if I just did accents without trying to too hard to do the age/sex - just getting the prosody right).

    I also tried AI-voice changing using commercial services boasting an array of character voices.

    And finally I hand-marked up characters and prosody on Google Play and got an excellent result.

    This GooglePlay version of Time for PsyQ is the best audiobook I've heard with its 20+ different AI-generated character voices - much better than the Apple and Amazon/Audible versions of my various books, and at least as good as any of the versions (of early chapters) that I produced myself using direct narration or voice munging. 

    It is available on Audiobooks, Kobo and Google Play:

    https://au.audiobooks.com/audiobook/time-for-psyq-kids/784375

    https://play.google.com/store/audiobooks/details/Marti_Ward_Time_for_PsyQ?id=AQAAAEDS3GicqM


    There are also AI-generated Apple and Amazon/Audible versions of all my books, and while they are pretty good and improving, they aren't nearly as good as the above Google Play version of Time for PsyQ. 

    Quantum Talents/PsyQ

    I've already written/blogged appendices to my Quantum Talent Series' Time for PsyQ (included in the education edition) so won't say any more about them here, but the educational version including the background material from the following blog entries has got some good reviews and feedback from parents and teachers:


    The GooglePlay version of Time for PsyQ is the best audiobook I've heard with its 20+ different AI-generated character voices - much better than the Apple and Amazon/Audible versions of my books. 

    The Educational Edition is here:
    https://books2read.com/Time4PsyQ

    (Both the educational edition with the Science for PsyQ appendix and the minimally priced original edition without the appendix are now on Amazon.)

    Do let me know what you think, and whether I should invest the time in doing Google Play versions of other audiobooks - I think they are better than anything I've heard on Apple or Amazon or Audible.



    Major outlets for Marti Ward books



    Contact Me


    I'm also open to other comments on my books or my blog, or requests to be a proofreader or betareader for my books (although I get lots of email, not to mention a lot of spam - so it may be a few days before I reply). I can be reached on gmail.com as marti24ward.