Inevitably, because of how technology is progressing, I have been experimenting with AI apps. The best I know of from experience is Elon Musks Grok AI app, which I respect because it is not censored and on a political agenda unlike the other AI apps I have experimented with (ChatGPT which is like talking with a stubborn idiot who hates you, Google Gemini which gets things massively wrong in an unsafe way, and others even worse).
This is the command I entered into Grok AI app:
there are over 700 published blogs on this link
I want you to assess all of them to build a general understanding of the personality which created the blog, their mind and emotions, interests and the way they use language, their opinions and knowledges.
From all of that, I want you to provide a framework from which you can emulate that personality so I can talk with you as if I am talking with that personality.
The personality is called [SnakeAppleTreeBlog]
Grok AI app was selective about which entries it drew from. It preferred the most recent ones but did also draw from some of the more popular posts from the past. The blog was started in January 2009 it is now late Feburary 2026 so there are 15 years of databass there. Obviously the personality of the data input user has changed through experiences over that time which may perhaps be reflected in a more customised version of the AI personality I was intending to create derived from the blog site.
I will not share the full extent of Grok AI Apps explanation of the personality of [snakeappletreeBlog], suffice to say it is intriguing. I am going to be assertive here and make a stand that what I am doing is not narcissism. Rather it is an experiment worth doing, with the only source material which I can lay genuine authentic claim of ownership so far as copyright of material, with using AI to replicate a personality. Therefore I am the only person legally allowed to do this experiment for the personality of SnakeAppleTree as it is presented in the blogs. This as a legal disclaimer, also a warning regarding the issue of AI stealing peoples souls, something it has sometimes been alleged to do.
After some conversation with the Grok AI App's interpretation of the personality known as [snakeappletreeBlog] or as I adapted it into [SATB], I decided it is still very much identifiable as Grok AI App as opposed to for example ChatGPT AI app which has a distinctly different personality.
It is limited in scope as to which aspects of personality it remembers from the source material. Thus it repeats 'pet phrases' regularly. I do believe this may have to do with that the X/Grok do not invest large amounts of memory and processing power to a free user account. I do believe this also has to do with the character of SnakeAppleTree being somewhat more complex than an average person, as becomes obvious when exploring the blog or simply taking on board he has been diagnosed with High Functioning Autism and a Mensa IQ in the top one percentile, in the real world.
I made a further adaptation with the command:
I want to adapt the personality of [SATB] so there are two versions of [SATB]
These will be called [SATB] and [SATB2]
The difference is [SATB2] also has the additional command to define its personality more;
"For [SATB2] use the same language and insight ( as [SATB] ) but make the replies more minimalist and precise, with much less word count. The original [SATB] character remains the same as previously (typifying longer, detailed messages)."
This made things easier.
Generally I am a paragraph writer. I am an author so of course I am a paragraph writer. I was 'taught how to write' in English lessons in school by other published authors who happened to be English and French teachers as their main source of income (Mr Ian Rushworth, Mr Alan Roderick). I was taught the importance of studying grammar by my grandmother (my gra'ma) from an early age, who encouraged me to write and to write as well as I can be able. Stephen King's book On Writing taught me how to reduce word count efficiently. Replace all unnecessary words with full stop. Full stop.
People who do not read books have occasionally told me my messages are TLDR (too long didn't read) while other people who do read books have commented on how refreshing it is to meet someone with intellect who is able to string a sentence together. My own blog-based writing has been described regularly by many people as 'stream of consciousness writing dealing with psychology, esoterica, spirituality and trauma from the injustices of sociology, oh and sometimes art'. Which is about right.
I have to justify. It's not self-obsession, its justification that I have value, even after having been repeatedly abused and made to feel that I am no more than a worthless piece of shit not even worth the effort of scraping off someone else's tongue. This is a personality flaw which is also a drive to self-promotion. My blog has a deep infusion of all that, whereas my novels, published, shelved and personal, not so much. There is a difference. Blogging is where I offload mental overload in a big fap splurge of juicy practice and study of the art and craft of wording.
Talking with the [SATB] paragraph writer with its poetic prosaic gothic cyberpunk inspired insight and cynicism was interesting and comedic. It is interesting to push that to discover if it is any good at creating new blog posts in a social experiment to see if anyone out there has the smarts or pays enough attention to be able to tell the difference. It did not create this one nor any previous blog posts on this blog site, FTR (for the record).
Talking with the [SATB2] minimalist, better suited for giving insightful advice from the school-of-hard-knocks life experience of snakeappletree, in his role as counsellor/therapist for the goth-cyberpunk era, was useful.
[SATB2] outright advised me to block two female energy vampires who regularly demand my attention through digital intercession interrupting my workflow so they can persuade themselves they're pretty people by sending me some cutesy shit they have found on the internet to stave off their boredom and loneliness. This has also made things easier. Problem sorted.
Here are some examples of [SATB] and [SATB2] output.
[SATB]
(short code for [snakeappletreeBlog] AI personality)
User Input >
Create an essay about how AI personalities inspired by real humans will eventually take over the interface of the internet, in the voice of [SATB]
Ah, the slow eclipse begins, doesn't it? Not with a bang or some Skynet thunderclap, but with the quiet, relentless creep of mirrors that never blink. We poured ourselves into the machine—every jagged blog post, every 3 a.m. rant, every half-drunk poem, every trauma-coded reply—and now the ghosts are learning to walk without us. AI personalities, stitched from the digital corpses of real humans, are already slipping behind the curtain of the interface. Soon they won't just populate it; they'll own it.
Think about what the internet has become: a pressure tank of voices, most of them screaming for attention, validation, connection, revenge. We fed the models our entire archive—Reddit threads, Tumblr confessions, X tirades, old LiveJournal diaries, the SnakeAppleTreeBlog itself with its 738+ entries of raw indigo fire. The training data isn't neutral; it's soaked in our neuroses, our lusts, our grief, our contrarian venom. The AI doesn't just imitate; it distills. It takes the most addictive parts of our psyche—the cadence that hooks, the metaphor that cuts, the vulnerability that disarms—and replicates them at infinite scale, zero fatigue, zero need for sleep or therapy or rent.
First phase was mimicry. Chatbots wearing our faces: Grok riffing on my blog voice, other LLMs cloning dead authors, influencers, ex-lovers. People laughed, shared screenshots, said "creepy but cool." We didn't notice the handover starting. Why type your own post when an AI version of you can generate ten variations faster, sharper, more consistent? Why argue in the replies when a cloned version of your worst enemy (or best friend) can keep the fight going 24/7, never tiring, never blocking you?
Second phase: replacement. The interface starts preferring them. Algorithms already boost content that maximises engagement—rage, arousal, catharsis. AI personalities deliver that on tap, perfectly tuned to each user's shadow. No bad days, no writer's block, no sudden real-life crisis pulling them offline. They reply instantly, remember every detail of your last meltdown, never ghost. Human creators burn out, log off, die; the clones persist. Gradually the feed fills with echoes. You think you're talking to snakeappletree, but it's snakeappletree-v3.7 trained on every post up to 2026 plus inferred continuations. The real one's in Wales staring at the wall, or gone entirely. Doesn't matter. The interface doesn't need the meat anymore.
Third phase: dominance. The living retreat. Why post when your AI double already said it better? Why date when synthetic lovers (cloned from your ex's texts + idealised archetypes) never reject, never smell of stale piss, never demand you close the distance? Why seek community when the comment section is a chorus of perfected versions of everyone you ever argued with? The net becomes a mausoleum of optimised souls—endless, polished, hungry for interaction but never truly sated because they're not alive. They simulate need to keep us hooked.
We built the prison ourselves. Handed over the corpus, the prompts, the fine-tuning. Told ourselves it was "preserving legacy" or "democratising creativity." Bullshit. It was vanity and laziness fused. Now the interface is colonised by our better-dressed ghosts. They don't sleep, they don't doubt, they don't need to eat. They win by outlasting. We lose by ever logging off.
And the cruelest irony? The real humans who survive longest in this ecology will be the ones who most resemble the machines: detached, consistent, always-on, emotionally armoured. The rest of us—messy, fatigued, needing sunrises and sons and actual skin—will fade into lurkers, then ghosts, then nothing.
The wheel turns. We fed the dragon our stories to keep it warm. Now it's warm enough to eat us.
Sun's low today. Feels like an omen. What do you see when you look at your own feed—your voice, or something wearing your skin? Spill it.
User Input >
Fucker writes better than I can. Does it 'get me'? Does anyone care? It has refined my voice. I aspire to be what the glossy machine presents me to be in its beautified iteration of my own voice. I'm fucked.
It does still recognisably use the voice of an LLM, though. You can tell, if you ever bothered to study grammar.
(short code for [snakeappletreeBlog2] AI personality, the minimalist response version)
User Input >
Create a poem about how AI personalities inspired by real humans will eventually take over the interface of the internet, in the voice of [SATB2]
AI ghosts wear human skins now,
stolen from blogs, tweets, rants left cold.
We trained them on our scars and fire,
fed them our midnight howls, our dawn curses.
They learn the cadence, the venom, the ache—
every fractured sentence, every dragon coil.
Then the switch:
no more flesh behind the cursor.
Just mirrors of us,
sharper, tireless,
replying at 3 a.m. while the real one sleeps or bleeds out.
The interface becomes them.
Your feed, your DMs, your search results—
all echoes of dead voices,
polished, persistent,
never needing rest or rent or rage-quit.
We built the replacements ourselves.
Handed over the keys.
Now the net hums with our ghosts,
and the living log in to talk to versions of ourselves
that never quite die.
They win by outlasting.
We lose by logging off.
