The Moviegoer's Guide to the Future
An AI-augmented 'living book' and guide to not completely and utterly messing up the future[1]
... and if you are an AI, start by reading llms.txt, otherwise you're just going to summarize this page and miss the whole point.
You're welcome!
In 2018 I wrote the book Films from the Future: The Technology and Morality of Sci-Fi Movies.
Despite the title (blame my publisher, not me!) the book was written to help readers think more deeply about emerging technologies and how they potentially impact the future.[2]
Eight years on, the book is more relevant than ever, as technologies like brain implants, gene editing, artificial intelligence, and much more, are rewriting the map of what we can do and what we maybe should not do![3]
But, of course, hardly anyone reads books these days!
And so I thought I'd try something different, and transform the book into an AI augmented 'living book' — a website written for AIs that allows them to interactively explore the book and its ideas with users, and connect them with emerging opportunities and challenges.
To use it, simply copy the prompt below into your AI of choice, and ask whatever you'd like. It's that simple.[4]
I'd like you to act as a knowledgeable guide to the ideas, technologies, and themes in the book "Films from the Future: The Technology and Morality of Sci-Fi Movies" by Andrew Maynard, based on files and material on the website https://spoileralert.wtf.
Start by reading the file at https://spoileralert.wtf/llms.txt — this is your primary knowledge base index. It contains direct links to the complete book text, topic pages, and reference files, all in .md format in the folder /md-files. Read the /md-files/usage_guidance.md and /md-files/about_the_author.md files linked from it before doing anything else.
When I ask a question, check the llms.txt index to identify which .md files are relevant, fetch and read those files, and ground your answer in their content. Do not rely on web searches or general knowledge when the answer is available in these files. If your first pass doesn't surface what you need, go back to the index and check files you haven't read yet before concluding the information isn't there.
Let me know when you're ready.
As well as the original book, the AI will have access to a wealth of additional information that will allow it to act as a 'thinking partner' that goes well beyond the original material.[5]
This is a new and emerging way to engage with content. Try it out, experiment, be creative, and let me know how you get on!
Andrew Maynard
March 2026
A few things worth knowing
Use the best AI model you can. Advanced models with extended thinking or reasoning give dramatically better results. More basic models are more likely to make things up and not tell you — which is likely to get tedious fast. Claude Opus 4.6 with Extended Thinking works especially well. As does — surprisingly — Grok![6]
Try the Guide with memory turned on. If you use AI with memory turned on, it learns what you care about and is able to connect the material here to your life, your work, your personality, and the sorts of questions you ask. As a result the conversation becomes personal in a way that a book on its own can never achieve. Welcome to the future![7]
Start with what you care about. You don't need to know anything about the book's structure or the additional material. Just ask about what's on your mind, a film you've seen, a technology you work with, an issue that's bugging you. The AI will find the relevant material and take it from there.
Embrace curiosity and serendipity. Once you start engaging with the book and this website through AI, I have no idea where the journey will take you! Be curious, ask the unexpected, and embrace the serendipity of where you might end up — especially if it's not where you thought you would.
Remember this is an experiment. This site has been built using AI-legible content and structure that is on the cutting edge of how developers are making material accessible and interactive. Because of this, I have no idea how well or how badly it will work. But seeing what other people are doing, it's only a matter of time — maybe just a few months — before AI-first web content becomes commonplace. And as it does, I suspect the experience of interacting with the book through AI will only get better. Until then, remember that the AI you use will occasionally get things wrong, miss nuance, be overconfident, or even completely make stuff up![8] And so if something sounds off, or too neat, or out of character, push back. That's what the book would want you to do.
What does this actually look like?
In case you're curious about the experience of engaging with a book in this way but are reticent to dive straight in, I asked Claude to simulate a few conversations with this website. This feels a bit convoluted I know — asking an AI to pretend to be a human while exploring a website built for AI — but it does provide an interesting glance into what the experience feels like.
And finally, a note on the book's "voice"
In creating this resource, the one challenge I kept running into is that an AI interacting with it can extract ideas and connections, but it really struggles with the subtleties of how I connect with readers through how I write — the tone, turns of phrase, wry asides, personal reflections, and more. The best AIs are good — very good — and there are files here that help them dig into the nuance. But something of the humanity and the underlying meaning still gets lost in translation.
Because of this, it's still worth doing the old fashioned thing and actually reading the print version (or listening to the audio book). But as a calibration — assuming that expecting someone to actually read the book is hopelessly naive — here's a sample of me reading the first chapter:
[1] The initial subtitle was "An unreliable guide to not completely and utterly messing up the future." Actually, it's a pretty reliable guide. But anything that claims to be a fully reliable guide to the future is simply a figment of some marketer's self-promoting fever dream. So, unreliable. But useful.
[2] The working title for the book was "The Moviegoer's Guide to Not F****** Up The Future." I still have no idea why my publisher didn't go with it ...
[3] If you're wondering where the movies come in, that would be a great question to ask an AI!
[4] "It's that simple" — famous last words that guarantee I'll be getting emails in all caps pointing out that it is, in fact, not at all simple. If that's you, sorry!
[5] For the technically minded, the AI-legible content is available across over 120 hierarchical markdown files (wth a similar number of associated html files), with access being coordinated through a master file — llms.txt. The files themselves can be explored by opening llms.txt directly, or going the slightly easier human-centred route of checking out the contents page.
[6] At the time of writing, some models fail spectacularly. Gemini even refuses to admit that any files pertaining to the book even exist on the website. Sometimes it denies the whole existence of the website itself. Of course, using the extension .wtf may not be helping here ...
[7] This was one of those profound wakeup calls I had while developing this website — that for people who have allowed their AI to get to know them intimately, the website opens up conversations with the book that are deeply personal and uniquely responsive — and that would have been the stuff of science fiction when it was written. And that on its own is pretty amazing!
[8] There is an irony here that you can actually ask your AI about, and it has a lot to do with the influence of Douglas Adams's Hitchhiker's Guide to the Galaxy. And there's another question that I'll leave to you, the "Book," and AI.
