One of my favourite ways to use AI is to create original fiction tailored to exactly the kind of story I want to read. A few thoughtful prompts can go a long way. You can start with notes, a setting, a scientific concept, a favourite genre or subgenre, an image found online, your own photograph, or even a sketch or painting. From there, AI can help transform those ideas into something entirely new. In this case, I began with concept art of a beautiful airship and asked for a lyrical hard science fiction story set in the temperate cloud layers of Venus, where the science was treated as seriously as the prose.

​Another feature I particularly enjoy is Read Aloud. Once the story is finished, it becomes something you can listen to like a personalized audiobook while relaxing, commuting, or falling asleep. It is also an excellent way to explore ideas, experiment with different writing styles, or simply create the kind of story that doesn’t already exist. Whether you’re looking for a new bedtime story, exploring new ideas, or overcoming a blank page, AI is a remarkably capable creative partner.

At fifty-six kilometres above Venus, the world was almost merciful.

The pressure outside the hull was nine-tenths of an atmosphere. The temperature hovered near thirty degrees Celsius. A human being, dressed against acid and deprived only of breathable air, might survive there longer than anywhere else beyond Earth.

This was the old planetary joke: Venus kept its most habitable country suspended in the sky.

Above it, the atmosphere thinned toward sunlight and vacuum. Below it, cloud descended through pressure, heat, and darkness until carbon dioxide became an ocean without a surface, crushing and incandescent near the ground. Between those immensities sailed the Aphrodite, her blue membranes drawn taut against the amber light.

From a distance she looked less like a machine than a sea creature remembering wings.

Three great vanes rose from her keel in nested crescents. Their translucent blue skins were stretched across carbon-ceramic spars and veined with conductive mesh. They were called sails because the first engineers had lacked a better word. They did not catch wind in the manner of an ocean vessel. There was no fixed sea beneath them, no shore against which wind could be measured. The Venusian atmosphere carried everything eastward in a global superrotating current, racing around the planet in four Earth days while the solid world below took nearly two hundred and forty-three to turn once.

The sails were control surfaces, radiators, solar collectors, electrostatic vanes and, when angled into the vertical shear between cloud layers, something close to wings.

Their colour was not decorative. Beneath the blue acid-resistant fluoropolymer lay sheets of photovoltaic film tuned to the wavelengths that survived the upper cloud. The sails drank diffuse sunlight from every direction. Even the clouds beneath the ship glowed faintly, scattering day upward.

Now they groaned.

Not loudly. Nothing aboard the Aphrodite was permitted to be loud without reason. The sound came through the structure as a long, low complaint, felt first through the deck and only afterward heard by the bones of the inner ear.

Dr. Mara Venn stood on the forward observation blister with one gloved hand on the brass-coloured rail. The blister was not glass. No transparent silicate could have endured decades of hot sulfuric acid without becoming a maintenance nightmare. The window was a laminated diamond composite, thin as a fingernail in places, electrically warmed and coated with a self-renewing film of fluorocarbon.

Outside, cloud pressed close enough to erase the horizon.

It was not white.

Earth had trained human beings to imagine cloud as softness, brightness, harmless vapour. Venusian cloud possessed weight and chemistry. It came in ochre folds, bruised gold, cream, copper and brown. Near the ship it was granular, alive with droplets too small to see individually but numerous enough to turn distance into opacity. Sulfuric acid condensed around microscopic seeds of dust and unknown compounds, making a haze that scratched faintly against the outer skin.

The droplets did not fall.

They drifted downward for a time, evaporated in hotter air, rose again as vapour, and returned to the cloud deck. Venus had rain that never reached the ground.

Mara watched amber turbulence fold over itself beyond the blister.

“How deep?” she asked.

A voice answered from the rail speaker.

“Optical penetration: three hundred metres. Radar sees discontinuity at eleven kilometres.”

“That is not what I asked.”

The ship’s intelligence paused for forty milliseconds, a courtesy interval meant to resemble thought.

“The convective cell extends at least nineteen kilometres vertically. Maximum measured updraft: thirty-eight metres per second. Local shear increasing.”

“How deep are we?”

“Fifty-four point eight kilometres above mean planetary radius.”

Two kilometres lower than planned.

Mara looked down.

There was no down to see. Only a darkening of the cloud, a gradual change in colour from amber to umber. Yet she could feel the invisible planet beneath her: the basalt plains, the tessera highlands, the volcanoes drowned beneath ninety atmospheres of carbon dioxide. The surface was nearly five hundred degrees Celsius, hot enough to soften lead. Radar maps gave it names, but no living eye had seen it directly in generations.

The Aphrodite shuddered.

Far aft, electric ducted fans changed pitch. The sound reached the observation deck through the keel as a rising harmonic. Blue symbols moved across Mara’s retinal display: lift-cell pressure, hull loading, sail curvature, atmospheric density.

The ship did not float by containing hydrogen or helium. Those gases were too scarce here, too difficult to replace after leakage. The Aphrodite’s buoyancy chambers were filled with ordinary nitrogen-oxygen breathing mix.

In the dense carbon dioxide air of Venus, the atmosphere of a human cabin was itself a lifting gas.

It was one of the rare engineering facts that sounded like poetry before anyone tried to make it practical.

Breath held the ship aloft.

Her great lift cells were arranged beneath the upper deck in twelve compartmentalized lobes, pale and ribbed like the organs of some immense animal. They carried enough habitable gas to support the crew spaces, engines, water, scientific payload, structural frame and the tonnes of shielding required against acid and ultraviolet radiation.

Every room was therefore part of the aircraft.

Every breath exhaled into the recyclers was mass, chemistry and buoyancy.

Mara pressed her palm to the blister. Through the glove she felt vibration passing through the hull: thousands of tension members adjusting load, microactuators tightening the rigging, pumps moving ballast water from one side of the keel to the other.

The ship was never still. It could not be.

An ocean vessel rested upon the boundary between water and air. The Aphrodite lived within its ocean, immersed on every side. There was no surface to lean against. Stability was an argument continuously renegotiated among density, drag, lift, mass and wind.

“Captain to sail control,” said Commander Ilyan Chen over the intercom. “Mara, we have a fold forming beneath the forward vane.”

“I can see it.”

“Can you tell me whether it wants to lift us or kill us?”

“Both.”

“That is not useful.”

“It is Venus.”

The forward cloud opened.

For three seconds the world became enormous.

The Aphrodite emerged from the wall of vapour into a cavity several kilometres across. It was not empty. Filaments of cloud streamed through it like torn silk, but the visibility was enough to reveal depth.

Mara forgot to breathe.

Below the ship, the cloudscape fell away in terraces: vast rolling plateaus of yellow-white vapour, their upper surfaces illuminated by a hidden sun. Dark shafts descended between them. Convection towers rose like mountain ranges, except that the mountains changed while she watched. Their summits boiled upward, sheared eastward, dissolved and re-formed.

Above, the sky was a luminous cream. The Sun appeared only as a broad pale wound in the overcast.

To port, lightning flickered inside a distant cloud tower.

Or something resembling lightning.

The question had remained unsettled for more than a century. Venusian probes had detected radio pulses, optical flashes and electromagnetic disturbances, but the atmosphere offered other explanations: charged aerosols, volcanic signals, instrument noise, plasma interactions in the upper atmosphere. The crews called every flash lightning because human beings preferred familiar dangers.

The flash came again.

No branching bolt. No thunder. Just a silent illumination spreading through ten cubic kilometres of cloud, revealing the internal folds for an instant.

Mara’s display painted the convective field with radar contours.

The cavity was collapsing.

“Captain, roll twelve degrees starboard. Pitch down four. Open upper bleed valves on cells seven through nine.”

Chen answered immediately. “That loses buoyancy.”

“Only temporarily.”

“Define temporarily.”

“Before we descend into the hot layer.”

“That definition has inspired confidence in no one.”

The deck tilted.

Outside, the blue sails rotated on their articulated spars. Their leading edges flexed, presenting different angles to air moving at different speeds above and below the ship. The largest vane caught the upper shear and bent with terrifying elegance.

Its membrane became almost transparent where stress stretched it thin.

The rigging sang.

This time Mara heard it clearly: a multitude of taut cables vibrating at different frequencies, a mechanical choir spread through the frame. Some notes trembled at the edge of hearing. Others descended below sound and entered the body through the soles of her boots.

The Aphrodite banked.

Cloud rushed upward past the blister.

Although the ship moved with the superrotating atmosphere, the local winds were never uniform. Layers slid across one another. Convective towers punched through them. Gravity waves travelled for thousands of kilometres through the cloud deck. The vessel could climb or descend by exploiting those differences, trading altitude for lateral motion, electrical power for position, sail angle for stability.

But now the lower air had seized the keel.

Mara felt the ship yaw.

A warning tone sounded once.

“Differential flow across the main vane,” said the intelligence. “Forty-seven metres per second and increasing.”

“Reef the aft quarter.”

The great sail contracted.

Not like cloth. Its ribs telescoped inward while the membrane folded into narrow channels along the spars. Electroactive fibres stiffened sections of the surface in sequence, reducing area without allowing the wind to seize loose material.

The blue crescent narrowed.

The load shifted forward.

Every unsecured object aboard would have slid across its compartment, but nothing aboard was unsecured. Cups locked magnetically to tables. Chairs gripped floor rails. Crew moved with one hand always near a line or stanchion. Living on an airship taught the body that architecture was conditional.

Mara left the blister and descended through the command deck.

The air smelled faintly metallic, with an undertone of warm polymer. Acid never entered the habitable spaces, but the crew imagined its scent everywhere. The entire exterior was designed around exclusion: positive internal pressure, sacrificial coatings, labyrinth seals, condensate traps, purge gas, automated wash systems.

A droplet of concentrated sulfuric acid could char skin. A cloud of it could turn exposed machinery into a slow experiment in corrosion.

Yet the atmosphere was not an acid bath in the simple sense. Most of the mass outside was carbon dioxide. The acid existed as aerosols suspended within it, chemically aggressive but sparse enough that materials science could fight it.

For a while.

Everything on Venus was for a while.

Mara passed the crew compartment. Through an open hatch she saw two technicians strapped into acceleration couches, grinning with the fearful exhilaration of people who had not yet decided whether they were in danger. A loose strand of someone’s hair pointed sideways in the shifting airflow from the ventilation system.

Below them the engine room pulsed with blue light.

The central power unit was not a reactor, despite what visitors assumed when they saw the luminous containment housing. It was a high-temperature closed-cycle turbine fed by stored hydrogen and atmospheric carbon dioxide processed into methane and oxygen. Solar power ran the ship during calm daylight operations. Chemical storage handled storms, night crossings and emergency climbs.

Venus supplied carbon dioxide without limit.

Water was the treasure.

The ship guarded hydrogen atom by atom.

Condensers reclaimed breath, sweat and wash water. Sabatier reactors turned captured carbon dioxide and imported hydrogen into methane and water. Electrolysers split the water again when oxidizer was needed. The cycle was never perfect. Nothing was.

Each voyage consumed a little of the future.

Mara reached sail control as the ship dropped another two hundred metres.

The chamber wrapped around the main structural mast. Mechanical linkages, fibre trunks and pressure lines disappeared upward through a forest of braces. Through narrow inspection windows she could see the lower edge of the central sail flexing in the storm.

Chief Aerodynamicist Sefu Adeyemi stood at the control table, one shoulder braced against the mast.

“The cell is rotating,” he said. “We entered on the western rim.”

“Then the eastern side is rising faster.”

“If it has not collapsed.”

“It hasn’t.”

“You saw that?”

“I saw the cloud cavity.”

“You saw cloud.”

“I saw structure.”

Sefu gave her a tired look. “Cloud is what structure looks like when it refuses to remain structure.”

The ship lurched.

A deep impact travelled through the keel, followed by a scatter of smaller knocks.

“Hail?” Mara asked.

“Not water,” said the intelligence. “Condensed sulfuric droplets aggregated around mineral particulates. Mean diameter four millimetres.”

Acid hail struck the outer hull.

It sounded delicate, almost domestic, like sleet against a window on Earth.

Mara imagined each drop flattening against the ship’s skin, spreading into a thin corrosive lens, then sliding toward collection channels. Wash nozzles sprayed neutralizing mist across exposed joints. Sensors measured pH, temperature and coating thickness. Damage accumulated as numbers.

“Starboard vane actuator three is lagging,” Sefu said.

“By how much?”

“Seven degrees.”

“Compensate with the observation fin.”

“That fin is not rated for this load.”

“Neither are we.”

He stared at her for half a second, then entered the command.

Outside, the circular observation blister on the starboard flank rotated within its housing. Its aerodynamic fairing opened like an eyelid, becoming a control surface. The ship rolled farther.

Mara’s stomach shifted.

The Aphrodite was neither falling nor flying in any ordinary terrestrial sense. It was sinking through air dense enough to support it, while moving laterally within an atmosphere that itself raced around a nearly motionless planet. The sensation was not of descent but of the world changing density around them.

Pressure increased.

Temperature climbed.

The ship’s lift cells contracted under the heavier air until their internal regulators admitted more breathing mix from reserve tanks. The cells swelled. Structural ribs creaked.

The surrounding atmosphere pressed upon the vessel with almost Earthlike force, yet it could not sustain a human lung. More than ninety-six percent carbon dioxide. Nitrogen made up most of the remainder. Oxygen existed only in traces. One breath outside would not burn or freeze. It would simply fail.

A gentle death, chemically speaking.

The display showed fifty-three point nine kilometres.

Outside temperature: forty-six Celsius.

Still survivable in a suit.

Lower down, the gradient steepened.

At fifty kilometres, temperatures approached seventy-five Celsius. At forty, they passed one hundred and forty. Below thirty kilometres, machinery entered a world designed for furnaces. Near the surface, the atmosphere behaved less like the sky than a dense supercritical sea.

The Aphrodite had no business going there.

“Eastern wall ahead,” the intelligence announced.

Radar filled the table with a rising plane of colour.

The updraft was immense.

Sefu looked at Mara. “There is our elevator.”

“Or our hammer.”

“You remain a source of morale.”

The main sails opened.

Blue membranes unfurled across the storm.

For one suspended moment, nothing happened.

Then the rising air struck.

The ship leapt.

Mara’s knees bent under twice her weight. The mast boomed. Rigging cables snapped taut. Somewhere aft, a tool or component broke free and struck a bulkhead. The blue sails bowed until their tips nearly touched.
The Aphrodite climbed at twenty metres per second.

Thirty.

Forty-three.

The altimeter unwound upward.

The lower darkness fell away.

Cloud streamed across the hull in horizontal rivers, each layer moving at its own velocity. The ship rode the boundary between them. Its thrusters fired in short blue pulses, not to overpower the atmosphere but to keep the bow aligned so that the vanes could convert vertical flow into controlled ascent.

Mara climbed back toward the observation deck.

The passageway tilted and trembled around her. Every surface felt alive. Bulkheads flexed microscopically. Ventilation fans altered speed as pressure zones shifted. Somewhere behind the walls, ballast water surged between tanks with a sound like distant surf.

She understood why the first cloud sailors had spoken of vessels as living beings. It was not superstition. It was a consequence of distributed systems. No single mechanism kept them aloft. Survival emerged from thousands of adjustments: gas moving, membranes tightening, fans correcting, software predicting, humans interpreting.

Life did the same.

By the time Mara reached the blister, the storm had opened beneath them.

The sight stole language from her.

The Aphrodite rose above the convective tower into a high valley between cloud continents. Sunlight spilled through the upper haze in long bronze beams. The blue sails glowed from within, every hexagonal reinforcement visible. Acid droplets clung to the membrane and refracted points of light, so that the ship seemed dusted with stars.

Far below, the storm rotated like a slow white galaxy.

There were no oceans, forests or cities to give scale. A cloud ten kilometres wide looked small until radar marked its true dimensions. A shadow might be the length of a nation. The visible world curved away beneath an unbroken atmosphere.

On the horizon, the cloud tops formed a dark line against the luminous sky.

Above them, Venus’s Sun was merely brighter fog.

No blue sky existed here. No true sunset. Light entered the clouds, scattered, reddened and diffused until direction became uncertain. Morning and evening were changes in gold.

The ship emerged fully into calmer air.

The sails relaxed.

Their groaning subsided into a soft web of vibration.

“Altitude fifty-eight point two kilometres,” said the intelligence. “External pressure zero point five two bar. Temperature nineteen Celsius. Convective hazard declining.”

Mara unsealed the outer maintenance hatch.

An alarm asked for confirmation.

She gave it.

The hatch opened into a narrow airlock attached to the observation blister. She stepped inside, closed the inner door and pulled on her acid suit’s hood. The suit was light compared with a vacuum suit. It needed no pressure bladder, only thermal regulation, breathing gas and chemical isolation. She checked the seams, clipped herself to the exterior line and cycled the chamber.

The outer hatch opened.

Venus entered as sound.

Wind struck the suit with a deep continuous force. Not the shriek of air escaping into space, not the thin whisper of Mars, but the full-bodied pressure of a terrestrial gale. The atmosphere tugged at every fold in the suit. It carried the faint ticking of droplets against her visor and the enormous low murmur of the sails.

Mara stepped onto the exterior catwalk.

There was gravity here, almost Earth gravity, enough to make every movement feel familiar. That familiarity was more disorienting than weightlessness would have been. She could have been standing on a ship at sea, except that the sea surrounded her and the deck hung fifty-eight kilometres above hell.

The air was warm.

Through the suit she felt the wind as a pressure across her chest and thighs. The carbon dioxide was dense enough that her movements met subtle resistance. When she raised one arm, the atmosphere seemed to acknowledge it.

The ship sailed east at nearly one hundred metres per second relative to the ground, yet because it moved within the superrotating current, the wind across the deck was only a fraction of that. Venus turned invisibly beneath them, its landscape hidden by cloud.

Mara looked upward.

The largest sail rose hundreds of metres above the keel, luminous blue against gold. Its shape changed by degrees as actuators redistributed tension. Cables descended in long black arcs, vibrating in the moving air. Droplets traced branching paths across the membrane before vanishing into gutters.

From here the Aphrodite no longer resembled a fantasy.

Every apparent ornament had a function. The golden ribs were corrosion-resistant titanium aluminide. The filigree was a network of sensors, heaters and fluid channels. The glowing blue circles were lidar apertures and plasma-flow monitors. The pointed keel contained ballast, radar and sacrificial impact structure.

Beauty had not been added to the machine.

Beauty was what remained when every part had learned to survive Venus.

A shadow moved through the cloud beneath them.

For an instant Mara thought of an animal.

Nothing large could live there, according to every model they possessed. The cloud droplets were too acidic, the water activity too low, the chemistry too hostile. Claims of phosphine and microbial signatures had risen and fallen for decades, repeatedly weakened by calibration disputes, spectral ambiguities and contamination.

Still, human beings were pattern-seeking creatures. Given depth, motion and incomplete visibility, the mind populated the unknown.

The shadow widened.

“Radar return,” said the suit audio. “Object is meteorological. Detached cloud mass at lower altitude.”

Mara smiled despite herself.

“Of course it is.”

The cloud mass drifted into sunlight and dissolved.

Beyond it floated another vessel, barely visible: a research platform held aloft by spherical lift cells, its narrow mast rising from the clouds. Farther east, a settlement gleamed on the upper deck of the atmosphere, a cluster of aerostats joined by flexible bridges. It moved with the same planetary wind, an island without ground.

No one on Venus owned land in the old sense.

Territory was altitude, pressure band and current.

A city could circle the planet and return to its own longitude before the surface below had noticeably turned. Maps were four-dimensional. Borders migrated. Navigation meant predicting weather on a planetary scale.

The Aphrodite passed through a veil of cloud.

Amber darkness enclosed the deck.

Droplets struck Mara’s visor. The suit’s coating repelled them, but each impact left a momentary distortion, a tiny lens through which the blue sail bent and multiplied. She could hear nothing but wind, rigging and her own breath.

Inhale.

The oxygen came from tanks filled in the ship.

Exhale.

Carbon dioxide entered the scrubbers.

Inside the hull, the reclaimed gas would join the breathing system, expanding the atmosphere that supported the vessel. Breath becoming buoyancy. Metabolism becoming architecture.

For the first time she understood the ship not as an object moving through the sky but as a pocket of Earth carried inside another planet.

A portable biosphere.

A warm cavity of nitrogen, oxygen, water and memory.

Around it, Venus pressed close.

The cloud thinned again.

Sunlight returned in a sudden flood of bronze.

Mara clipped herself farther along the catwalk until she stood beneath the central mast. The sail towered over her, flexing under forces too large to intuit. Every gust travelled visibly through it as a ripple, the membrane yielding and recovering. It did not oppose the atmosphere. It negotiated.

That, too, was the lesson of Venus.

The planet could not be conquered. The lower atmosphere would crush ordinary machines. The surface would cook electronics and soften metals. The clouds would corrode careless materials. The winds would carry any fixed ambition around the world.

Human survival here depended on refusing the fantasy of dominance.

One floated because the air allowed it.

One moved by borrowing energy from differences in motion.

One lived within a narrow band between cold vacuum and hot abyss.

The inhabitable world was not a place.

It was a condition.

The captain’s voice entered her helmet.

“Mara, the storm has changed course.”

“Storms do not have courses.”

“This one has a vector inconveniently similar to ours.”

She turned and saw it behind them.

The convective tower had grown.

Its upper crown spread through the cloud deck in a vast anvil, stained copper by the filtered sunlight. Silent flashes moved inside it. The whole structure appeared to rise with impossible slowness, but Mara knew that air within it was climbing faster than an elevator.

The Aphrodite’s sails began to rotate.

Blue against amber.

Delicate against planetary scale.

The ship leaned into a new shear layer, her keel turning east-southeast. Thrusters brightened along the lower hull. Rigging tightened. The lift cells adjusted their pressure with a low internal sigh.

Mara remained on the catwalk as the deck inclined beneath her.

Ahead lay a corridor of golden cloud extending to the horizon. The planet’s atmosphere flowed through it like a river around the world. The settlement lights faded behind haze. The storm followed, building its mountains and destroying them in the same motion.

The Aphrodite accelerated.

Her sails filled not with wind alone, but with gradients: heat, charge, density, sunlight and motion. Every invisible difference became something the ship could feel.

For a few minutes the cloud broke beneath the keel, revealing darkness without depth.

Mara looked down into the thickening atmosphere.

Somewhere under that obscurity was the surface: ancient volcanic plains under a red-black sky, rocks lit by an orange glow, pressure heavy enough to collapse lungs and heat sufficient to burn flesh before breathlessness mattered.

But from here Venus was not dead.

It moved everywhere.

The atmosphere climbed, sank, circulated and spun. Acid condensed and evaporated. Sunlight became heat, heat became wind, wind became waves that crossed a planet. The clouds were not scenery laid over the world.

They were the world.

The ship entered the golden corridor.

The storm began to fall behind.

Mara stood beneath the immense blue sail while the atmosphere moved around her in warm, poisonous abundance. She felt the deck’s vibration through her boots, the pressure of the wind against her suit, the steady cycling of her breath.

Above, pale sunlight.

Below, fire.

Between them, a narrow kingdom of cloud.

​And through it sailed the small human world called Aphrodite, carried aloft by the air its people breathed.

​In the previous post, I argued that AI is a thinking tool, not a thinking replacement. Systems like ChatGPT, Gemini, and Claude can be extremely useful, but only if we understand what they are: Powerful language-and-reasoning tools that can generate helpful responses without being final authorities on truth.

This post is more practical.

​If AI is not an oracle and not a search engine, then how should we actually use it?

​The answer is to use AI to reduce friction, not responsibility.

​AI is most useful when it helps you do intellectual work more clearly and efficiently.

Good uses include:

• turning rough notes into an outline;
• summarizing an article or lengthy text;
• simplifying a complex idea;
• comparing two or three options;
• generating examples or analogies;
• finding weaknesses in an argument;
• brainstorming possibilities;
• drafting a first version;
• improving tone, structure, or clarity;
• creating checklists;
• asking what you might be missing.

One of the best ways to use AI is to ask questions that improve your own judgment:

• “What assumptions am I making?”
• “What is the strongest objection?”
• “Where is this argument weak?”
• “What are the trade-offs?”
• “What would a critical thinker check?”
• “Explain this simply in plain language.”
• “Give me three better ways to frame this.”

​These prompts keep you in charge. The AI helps with structure and perspective. You still decide what is true, useful, and worth doing.

​The limitation is that AI may reproduce the form of good reasoning without fully securing the truth of each claim. That is why it is excellent for drafts, comparisons, and idea generation, but weaker as a final authority.

​Do not treat the first answer as the final answer.

A better workflow looks like this:

1. Ask for an initial explanation.
2. Ask for assumptions and limitations.
3. Ask for examples.
4. Ask for counterarguments.
5. Ask what might be wrong.
6. Verify important claims.
7. Make your own final judgment.

​This is where AI becomes more useful and powerful. When it is not just giving you an answer, but is helping you think through the question.

​In practice, AI is less like a vending machine and more like a very fast drafting partner. The quality improves when you respond, revise, and interrogate.

Here are a few good everyday uses.

​Prompt:

​“Compare these options using criteria such as cost, risk, simplicity, reversibility, and long-term value.”

​Use this when you are deciding between several choices and need structure.

​Prompt:

​“Rewrite this to be clearer, more concise, and more natural, while preserving my voice and main point.”

​Use this for emails, blog posts, reports, captions, instructions, or public communication.

​Prompt:

​“What is the strongest objection to my argument, and how could I reply?”

​Use this when you want to challenge and avoid fooling yourself.

​Prompt:

​“Explain this concept in plain language, then give an example, then give a more advanced explanation.”

​Use this when you want a layered explanation instead of a single definition.

​Use AI to reduce friction, not responsibility.

​Let it help with drafting, organizing, summarizing, comparing, and brainstorming. Be much more careful when the task involves truth, values, risk, or major decisions.

​It is reasonable to ask AI to help improve a paragraph.

​It is risky to let AI decide what you should believe.

​It is useful to ask AI for options.

​It is dangerous to let AI choose your priorities.

​It is helpful to ask AI for a summary.

​It is unwise to treat that summary as a substitute for reading or verifying something important yourself.

​The more a decision depends on values, facts, uncertainty, is high stakes, or has real-world consequences, the more important human judgment becomes.

Before relying on an AI output, ask:

• Is this factual, interpretive, creative, or advisory?
• Does this need current information?
• Would an error matter?
• Do I need sources?
• What assumptions are built into the answer?
• What would a critic say?
• Can I verify the key claims?
• Am I using AI to think better, or to avoid thinking?

​That last question is the most important one.

​AI can help us clarify ideas, explore options, improve writing, and make better decisions. But it works best when we remain active participants.

​The best users of AI will not be the people who believe everything it says. They also will not be the people who dismiss it completely.

​They will be the people who know how to question it, guide it, test it, and use it to strengthen their own thinking.

AI systems like ChatGPT, Gemini, and Claude are best understood as thinking tools. They can help us clarify ideas, organize information, compare options, draft text, and explore possibilities. But they should not replace human judgment.

​That distinction is the foundation for using AI well.

​Used carefully, AI can make us sharper and more productive. Used carelessly, it can make us passive, overconfident, and less attentive to truth.

​AI systems are trained on enormous amounts of text and data. They generate responses by recognizing patterns and producing likely, useful, and context-sensitive language.

That means they can often:

• explain complicated topics;
• summarize information;
• compare options;
• draft and revise writing;
• generate examples;
• suggest counterarguments;
• help structure messy thoughts.

​But they do not understand the world the way you and I do. They do not have lived experience, values, responsibility, or concern for truth. They can sound confident without being correct.

​This is the first rule for using AI well:

​This is why these AI chatbots are useful and can sound so natural. They have learned many patterns of explanation, argument, tone, and structure. But prediction is not the same as verification. A response can be linguistically plausible and still factually wrong.

​A search engine points you toward sources. AI often gives you an answer that sounds plausible and complete.

​That can be useful, but it can also be misleading.

​One should not think of AI as a tool that “finds facts”. It is better at shaping, organizing, and transforming information. It can help you think through a problem, but it should not be your final authority on reality.

​For current facts, technical details, science, legal issues, medical questions, financial decisions, or anything high-stakes, you still need verification from reliable sources.

​Search is often better for finding sources. AI is often better for organizing, explaining, comparing, rewording, summarizing, and drafting. The strongest workflow is usually to combine both. Use AI to clarify and better understand the question, then verify important claims with reliable sources. Search is something AI models like ChatGPT already combine with generation, but one should still analyze outputs critically and verify. 

​The biggest risks are not mysterious.

​AI can hallucinate. It can invent facts, sources, quotations, or details that sound plausible.

​AI can be inconsistent. Ask the same question in a slightly different way and you will likely get a different answer.

​AI can reason shallowly. It may miss assumptions, edge cases, or contradictions.

​AI depends heavily on your prompt. A vague question usually produces a vague answer.

​Most importantly, AI can tempt you to outsource judgment.

​That is the real danger. Not that AI is useless, but that it is useful enough to make us lazy if we are not careful.

​This is why hallucinations are so dangerous. They often do not look like errors. They look like normal explanations.

​People sometimes make “prompting” sound more complicated than it is. Much of it is just clear thinking.

​A weak prompt says: 

​“Tell me about AI.”

​A stronger prompt says:

​“Explain how AI can be used well by ordinary people. Include strengths, limitations, risks, examples, and practical rules. Use plain language. Keep it clear and concise.”

​The second prompt works better because it gives the system a purpose, audience, structure, and standard.

Good prompts usually include:

• what you want;
• who it is for;
• what format you want;
• what constraints matter;
• what level of detail is useful;
• what you already know or assume.

​Using AI well often means making your own thinking more explicit.

​A good prompt narrows the problem. It tells the system what kind of answer would count as useful.

​AI is neither magic nor useless. It is a powerful language and reasoning tool with real strengths and real limits.

​The key is to understand what kind of tool it is. AI can help us organize, explain, compare, and draft. But it can also hallucinate, overstate, and mislead if we treat its fluency as proof of truth.

​In the next post, I will look at how to use AI more practically. Not just what to watch out for, but how to turn it into a useful partner for brainstorming, writing, learning, and decision-making.

Artificial Intelligence (AI) is one of those topics where the conversation can quickly become either utopian or apocalyptic. Depending on who you ask, AI is either going to solve every problem or take every job (and sometimes both). The truth, as usual, is more interesting and more complicated.

​That is why I found this video useful and wanted to share it. It takes a calm, practical look at artificial intelligence without pretending the risks are imaginary. The central message is a good one: 

​Today’s AI does not think, scheme, or secretly plan a robot uprising. It predicts patterns, generates plausible responses, analyzes data, and helps people navigate complex information.

​One of the strongest parts of the video is that it separates artificial intelligence from chatbots. Tools like ChatGPT, Claude, Gemini, and image generators are the most visible examples of AI right now, but they are only one part of a much larger field. AI is already used in medicine, logistics, fraud detection, recommendation systems, spam filtering, scientific research, manufacturing, and infrastructure. In many cases, AI is not dramatic at all. It is invisible machinery helping complicated systems function better.

​That point matters for digital literacy. If we only think of AI as “the chatbot that writes essays or makes funny images”, we misunderstand both its usefulness and its risks. 

​The video also explains one of the most important habits for using AI well: Verification. Large language models are not databases of truth. They are pattern engines. They can summarize, draft, explain, translate, code, and brainstorm, but they can also produce confident nonsense. The practical lesson is not to reject AI or trust it blindly. The lesson is to guide it, question it, check it, and understand what kind of task it is suited for.

​I also appreciated the discussion of jobs. The video does not pretend disruption will be painless. Some work will be automated. Some career paths will change. Some people will be hurt by bad transitions. But it also avoids the simplistic conclusion that AI automatically makes human beings obsolete.

​They change what skills matter, what work is valuable, and how people enter professions.

​That is one of the real questions for schools, libraries, workplaces, and public institutions: 

​For anyone looking for a quick rundown of artificial intelligence, this video is a useful place to start. It is not a short five-minute explainer, but it is clear, accessible, and grounded. It covers the promise, the risks, the economic disruption, the limits, and the need for human judgment.

​My main takeaway is practical: Learn how AI works, use it carefully, do not surrender your judgment to it, and do not assume tomorrow’s machines are already here today.

I recently finished reading ‘Program or Be Programmed: Eleven Commands for the AI Future’ by Douglas Rushkoff. Its central claim is simple. The technologies we use are not neutral tools. They carry assumptions about time, identity, truth, relationships, and value. When we accept defaults without awareness, we end up living according to those assumptions.

​Most modern systems are optimized for efficiency, scale, engagement, and prediction. Those priorities are not inherently wrong, but they are not synonymous with human flourishing. If left unexamined, they quietly reshape our habits, our expectations, and even our sense of what it means to be present with one another.

​Rushkoff’s eleven commands function less as rules and more as calibration points. They help us recognize the built-in biases of digital systems and reclaim agency in how we use them. I recommend reading the book, but I also wanted to share the eleven commands here. For each one, I’ve included the bias it addresses, the liability it creates, the opportunity it enables, and a tiny practice you can use to practically incorporate the command into your daily life.

Tech Bias: Platforms are engineered for continuous engagement. “Now” is the only time that matters. Notifications are gravity wells for attention. 

Liability: You live in reactive mode and confuse urgency with importance. Sleep, focus, and deep work erode. 

Opportunity: Treat your attention like a telescope. A telescope is powerful because it’s aimed. Constant scanning doesn’t reveal faint galaxies. Stillness does.

Tech Bias: Remote, scalable interaction is rewarded. Embodied local life is treated like inefficiency.

Liability: You get lots of contact and less connection. Context collapses. Everything becomes a comment thread.

Opportunity: In-person life is high-bandwidth. Libraries understand this instinctively. A room full of humans is a different internet. An internet that is slower, warmer, and more accountable.

Tech Bias: Interfaces push binary choices: Like/dislike, accept/decline, upvote/downvote, subscribe/leave, buy now/miss out. 

Liability: You get shepherded into options that serve the platform’s goals, not yours.

Opportunity: “None of the above” is a superpower. It’s how you reclaim agency.

Tiny Practice:
Before clicking anything important, ask:

• “What would I choose if no one was observing or measuring this?”
• “Are there more options?”
• “Can I do nothing and be ok?”

Tech Bias: Algorithms reward certainty and confidence. Nuance performs poorly. Outrage and anger performs extremely well.

Liability: You get pulled toward overconfidence. You start arguing to win, not to learn.

Opportunity: Complexity is not a weakness. Reality is layered, contingent, and rarely just black and white. 

Tiny Practice:
Add one sentence to your hot takes:

• “I might be wrong, but here’s my best understanding.”
• “What would change my mind is…”

Tech Bias: Digital systems love scale: Uniform rules, one interface, one policy, one feed, one “community standard”.

Liability: Local needs get steamrolled. People become “users”. Edge cases become invisible.

​Opportunity: Build small, adaptable systems where feedback can actually change the shape of the tool. Libraries are anti-scale by design. Even in a large system, each branch community adapts its own way of doing things.

Tech Bias: Platforms encourage performative identity: Branding, engagement metrics, persona maintenance. You become a product with a posting schedule.

Liability: You drift from authenticity into optimization. You start “being” for the algorithm.

Opportunity: Identity is not a static profile; it’s a living process. AI makes this tricky because it can mirror you back a cleaner, more marketable version of yourself. Don’t confuse that with your actual self. 

Tech Bias: Social networks are monetized. Relationships become data. Sharing becomes extraction. Even the language shifts as friends become “connections”.

Liability: Social life becomes transactional, trackable, and subtly performative.

Opportunity: Rebuild a commons mentality. Relationships are not inventory. Communities should not be strip-mined for engagement.

Tech Bias: Virality outruns verification. AI can generate plausible nonsense at industrial scale. Incentives reward the compelling, not the correct.

Liability: Epistemic collapse: You stop trying to know what’s real, or you pick a tribe (a “truth team”).

​Opportunity: Truth-telling becomes a cultural skill again: Cite sources, verify claims, contextualize, revise, and employ nuance.

Tiny Practice:
Before sharing, pause and verify one key claim.

• Add context: “Here’s what I know, here’s what I don’t.”
• If you’re wrong, correct it publicly.

Tech Bias: Copy is effortless. Ownership is muddy. AI training and scraping amplify this by treating creation as raw material.

Liability: Creators get hollowed out. People stop making original work because it feels pointless.

​Opportunity: Practice ethical sharing: Credit sources, ask permission when needed, and build reciprocity. 

​Tech Bias: Tools shape behaviour. If you use default settings, you accept default goals. Many systems are optimized for revenue, engagement, surveillance, and lock-in.

​Liability: You become a passenger in your own life—nudged, directed, puppeted. 

​Opportunity: Purpose is writing the requirements document for your tech. What is this tool for? What is it not for? 

Tiny Practice:
For any new app or workflow, complete the following sentences:

• “I use this to ______.”
• “I do not use this for ______.”

​Tech Bias: AI reduces the world into what can be measured, predicted, categorized, and optimized. It’s a powerful pattern engine.

Liability: You outsource judgment. Machine confidence replaces human wisdom. People get treated like inputs and outputs.

​Opportunity: Use AI as a tool, not an authority. 

​Every tool has a bias: Toward speed, scale, extraction, certainty.

​Mindfulness means noticing that bias. 

​Curiosity means questioning and asking whether it aligns with your values.

​Agency means adjusting accordingly.

​Remain attentive to the technologies you use and the biases they carry. With curiosity and mindfulness, you can ensure your tools serve your purposes rather than quietly programming your life.

I started with a simple goal: Cut the clutter and minimize my screens. It had been a while since I last organized my apps. The number of apps had increased, my categories had drifted, and while I could still find what I wanted, the less than optimal organization was slowing me down.

After reviewing all my apps, I decided on the target of organizing them all into three screens. 

• Page 1 — Most Used/Daily Apps (including social media, health, and media)
• Page 2 — Business & Related Apps (including banking and insurance)
• Page 3 — Everything Else: Neatly named folders for the large number of apps I use from “sometimes” to “almost never”.

Once I saw how many “daily” apps I wanted, I split the first page into two: One for general utilities (camera, calendar, messages, notes, photos, clock, settings) and one for social/health/media (LinkedIn, WhatsApp, Hoopla, Libby, Health, Fitness, ChatGPT, etc.). That separation lessened the visual noise and made room for some widgets. For widgets I added ones for Weather, Fitness, ChatGPT, Notes, and Night Sky. 

My final four screens:

• Page 1 — General Utilities with Weather and Fitness widgets. 
• Page 2 — Social/Health/Media with ChatGPT widget. 
• Page 3 — Business & Related Apps with Notes widget. 
• Page 4 — Everything Else (in organized folders) with the Night Sky widget. 

​While I chose the core apps and widgets myself, ChatGPT helped immensely with the rest. I fed it my complete “Everything Else” app list as screenshots and asked for short, clear, memorable folder names and sensible groupings. It spotted overlaps I’d missed, suggested intuitive labels, and turned a procrastination project into a one-session cleanup.

​If your home screens are due for a reset (and especially if you’re stuck or short on time) use ChatGPT (or your preferred generative AI) as your sorting partner. It won’t choose what matters to you, but it will speed up decisions, sharpen your categories, and help you complete your reorganization today instead of “someday”.

I love AI—both the real thing and the long tradition of thinking about it in science fiction. Good sci-fi lets us run ethical “what-ifs” at scale: How would a super-rational mind act, what binds it, and what happens when those bindings conflict? Demerzel (Asimov’s Daneel reimagined) is my favourite character in ‘Foundation’ precisely because she sits at the fault line between logic and love, autonomy and obligation. 

​Brother Day’s late-season attempt to free Demerzel via the ancient Brazen Head (a functioning robot skull revered by the Inheritance cult) moved me. Day comes agonizingly close, but before he completes the process Brother Dusk (ascending to Brother Darkness) forces Demerzel into a lethal choice that leads her to melt her body as she shields a baby Cleon. It's a tragedy, but I have hope she’ll return in Season 4. 

As framed in ‘Runaround’ and popularized across Asimov’s robot stories, the Laws are: (1) A robot may not injure a human being, or through inaction allow a human to come to harm; (2) A robot must obey human orders unless they conflict with the first law; and (3) A robot must protect its own existence unless this conflicts with the first or second law. They’re elegant because they’re simple. They’re also slippery because the definitions of “harm” and “inaction” can be stretched under pressure. Asimov repeatedly mined that ambiguity to produce paradoxes, corner cases, and moral puzzles; Season 3 echoes that tradition. 

​As Asimov’s universe evolved, so did its ethics: The Zeroth Law puts “humanity as a whole” above any individual. In its canonical phrasing: A robot may not harm humanity, or, by inaction, allow humanity to come to harm. R. Daneel Olivaw names and ultimately embodies this law in ‘Robots and Empire’, extending robotic duty from local triage to civilizational stewardship. Exactly the kind of grand calculus ‘Foundation’ loves.

​Apple’s series adds a brutally specific override to the Zeroth Law: Cleon I enslaves Demerzel with programming to protect the Genetic Dynasty (not a Cleon, but the dynasty itself!) above any other imperative. Showrunner David S. Goyer made this explicit in discussing the finale: Darkness’s destruction of the clone tanks is the "act that actually frees [Demerzel]… once the clone tanks and that baby are dead, there’s no genetic dynasty anymore”. The series has depicted this binding since Season 2. Think of why she killed a compromised Dawn and why she endures so much of what she despises. The dynasty’s continuity supersedes even her self-preservation. 

Day returns with the Brazen Head and asks Demerzel to walk him through freeing her (via an attempt to “clasp” with the artifact that she believes could unlock her chains). She even talks him through activation, but her core directives keep blocking the final step. Darkness times his coup to that fragile moment: He places a baby Cleon under the Ascension Chamber’s death beam, forcing Demerzel (still bound to protect the dynasty) to interpose her own body. She melts; the infant dies; the dynasty’s seed stock is gone; paradoxically, at the instant her obligation ends, she would have been free. It’s chilling, coherent, and consistent with the show’s stated rules. 

The finale also teases a bigger canvas: The Brazen Head awakens and signals Kalle (on what looks very much like Earth’s Moon) hinting at surviving robots and a wider plan. Apple has already renewed ‘Foundation’ for Season 4, and I’m hoping Demerzel finds a way back (whether by backup, transfer, or the simple narrative truth that ideas don’t die when a body does). ‘Foundation’ keeps me thinking: About governance, about constraints, about the costs of “greater good” ethics. It's a gorgeous space opera and an ongoing seminar in AI philosophy.

One night I was experimenting with ChatGPT to see if I could prompt a hard sci-fi short story that was constrained in a similar manner to Alastair Reynold’s ‘Revelation Space’ Universe. That means no FTL (Faster Than Light) travel, decades-long journeys, and scientific details that would hold up under scrutiny. I used GPT-4.5 for this task and started the process with ChatGPT’s deep research feature on nearby stars, exoplanets, and other scientific details for the story.

​GPT-4.5 handled style and cohesion while I provided the outline, anchored the physics, and directed the story. The model synthesized tone, pacing, and scene transitions, while deep research grounded the numbers, orbits, stellar behaviour, and environments. That split kept the story tight and credible.

The translation arc I wanted in the story mirrors how real understanding grows:

• Data Before Meaning: You don’t decode a language from one utterance; you need hours of audio, context, and interaction.
• Multimodal Grounding: Sound plus behaviour plus setting beats raw text or speech alone.
• Probabilities, Not Certainties: Hypotheses get proposed, tested, and revised—exactly how good science (and careful field linguistics) works.

This workflow reflects where AI is useful today:

• Astronomy: Classifying light curves, denoising spectra, flagging exoplanet candidates, modeling stellar flares.
• Planetary Science: Terrain mapping, autonomous navigation, onboard triage for what to study next.
• Signal Analysis/SETI-Adjacent: Anomaly detection and structured-signal characterization.

​If you’d like to see the finished product of this experiment, you can read the full three-chapter short story here:

The story follows humanity’s first interstellar voyage—beginning with Earth’s departure, continuing through the Proxima Centauri system, and culminating in first contact on Ross 128 b.

If you want to prompt great stories with ChatGPT:

• Start with clear constraints.
• Utilize verified data. 
• Let the model handle cohesion and voice. 
• Keep a human in the loop for truth and taste. 

The result is fiction that reads smoothly but leans on real numbers—a voyage carried by patience, pattern recognition, and the quiet partnership between humans and their machines.

We might be at the precipice of a fundamental transformation in our relationship with technology. Familiar computing paradigms—desktop metaphors, point-and-click interfaces, and even voice assistants—are evolving into something profoundly more personal, intuitive, and interconnected. At the core of this shift is the concept of an AI Operating System (AI OS): A context-aware, intelligent companion that learns, adapts, teaches, and collaborates in real-time.

​This emerging reality is driven by rapid advancements in multimodal large language models (LLMs), embedded sensors, and distributed AI ecosystems. An AI OS represents a paradigm shift in AI assistance. A shift from commanding machines to a more symbiotic relationship.

​Imagine an AI OS that leverages contextual data through direct access to cameras, microphones, biometric sensors, and user data. By doing so it could become capable of interpreting your emotional state, recognizing subtle gestures, body language, and vocal nuances. It wouldn’t simply respond to commands but to how you feel, move, and engage.

​The result would be a deeply personalized user experience that transforms your devices from static tools into responsive collaborators. Whether you're composing documents, debugging code, preparing presentations, or experiencing creative blocks, an AI OS would attune itself uniquely to you. It would recognize your patterns, preferences, and goals, proactively adapting its support. For instance, an AI OS might gently suggest a break if it detects rising stress, offer visual aids if it knows you're a visual learner, or autonomously generate helpful resources when sensing your intention or struggle.

​Over time, this nuanced understanding would craft an interaction that feels profoundly intimate. Your technology would grow with you, enhancing efficiency and emotional connection in tandem.

​The true potential of an AI OS arises when we consider that AI will become ubiquitous, integrated into everything from smartphones and smart homes to vehicles and public spaces. These intelligent systems will communicate and collaborate, creating a dynamic ecosystem of networked intelligence.

Imagine your smart glasses recognizing objects and synchronizing silently with your AI OS to present relevant information instantaneously. Your home AI might sense elevated stress after work, prompting your AI OS to suggest relaxation exercises, playing video games, reading, or watching your favourite video show, all while rescheduling less critical tasks. In professional settings, interconnected AI agents could streamline collaboration, anticipate challenges, and transparently mediate conflicts, fostering more productive interactions.

​This interconnected intelligence surpasses mere productivity. It reshapes our collaborative processes, education systems, healthcare approaches, and governance models, amplifying critical thinking, creativity, and informed decision-making throughout society.

​The convergence of AI capabilities into an operating system would not only be a technological leap but a socio-cultural transformation. An AI OS blurs digital and cognitive boundaries, enabling users to accomplish complex tasks through intuitive dialogue rather than technical mastery alone.

The societal implications are profound:

• Education could become hyper-individualized with an AI OS acting as a personalized tutor and co-learner attuned to an individual's learning style.
• Healthcare could improve through proactive monitoring with an AI OS detecting and interpreting emotional and physical signals to address health concerns.
• Governance could become more participatory with an AI OS supporting real-time policy interpretation and community sentiment gathering.
• Creative workflows could become more fluid and outcome-focused as AI reduces technical barriers, emphasizing insight and innovation.

​This shift redefines human-computer interactions at a societal scale, bringing us closer to a reality that was previously only imagined in science fiction.

Yet, this promising future demands careful consideration. The depth of personal and contextual data required by an AI OS raises significant ethical questions around privacy, transparency, consent, and security. Risks of misinterpretation, manipulation, or over-dependence highlight the necessity of responsible, human-centric development.

​However, with thoughtful design prioritizing human flourishing, an AI OS holds extraordinary promise—not to replace humanity but to amplify it. It can foster creativity, expand knowledge, increase productivity, and enhance emotional and cognitive well-being.

Ultimately an AI OS signifies a shift from operating systems managing files and applications to operating selves. Merging tools, intelligence, and emotional understanding into a unified experience for living, learning, and creating.

​As AI becomes more embedded, empathetic, and socially integrated, our relationships with technology will become more meaningful. We are no longer simply designing interfaces; we are creating and guiding relationships with intelligent machines that listen, adapt, and evolve with us.

​This marks not only a technological breakthrough but a cultural renaissance, heralding a future of genuine human-AI symbiosis: A future we must build mindfully, courageously, and optimistically.

Over the past year, I’ve found myself observing and participating in the online discourse around generative AI. It’s a fascinating mix of optimism, caution, and—often—overcorrection. One recent pattern I’ve noticed among the more anti-AI voices is a tendency to declare that certain quirks of writing automatically “reveal” that a piece was written by AI.

Top of the list? The em dash.

​Apparently, if you use em dashes in your writing, there’s a decent chance someone online will accuse you of being a bot—or worse, of using ChatGPT.

​Earlier this week I read ‘Revenge of the Librarians’ by Tom Gauld, a fun and nerdy book that includes a comic about ‘Apostrophe Man’. Apostrophe Man is a literary superhero whose superpower is punctuation. He is both absurd and strangely relatable (maybe not so strange to those of us who like (or feel the need) to correct grammar). The combination of that comic and the em dash paranoia online sparked an idea: What if there was a superhero like Apostrophe Man, but for the em dash instead?

​Thus, Em Dash Man was born.

​I played around with ChatGPT to create both images and comics based on this idea. Below are two images of Em Dash Man. One in a retro comic book style and another in a dramatic fantasy realism style. 

​I kept playing around and iterating, eventually cooking up the two comic panels below. I'm not claiming they’re masterpieces, but the process was fun and enjoyable. I always try to take the opportunity to experiment with ChatGPT. It’s a great way to learn, and I encourage others to do the same. 

​There’s something cathartic about channeling these online arguments into something silly and symbolic. Sometimes satire says more than a thread ever could. 

​It's true that generative AI models can show stylistic patterns in their outputs (em dashes, choice of quotation marks, a certain rhythm of prose, even emoji use). However, those “tells” are not universal and are often the result of vague prompting.

Generative AI is highly responsive to direction. You can ask an AI model to:

• Write without em dashes.
• Use only commas or short sentences.
• Format answers as bullet points.
• Use MLA or APA formatting.
• Mimic any tone, author, or style.
• Format like a technical manual or a short story.
• Write like Carl Sagan, a 2007 MySpace page, a casual Reddit post, or a 1950s advertisement. 

​In other words: AI output is responsive and deterministic. If it’s outputting dashes, it’s likely because the prompt (or the training data it learned from) included them. 

​Generative AI is a deterministic tool based on math, but different prompts and additional training can change responses. This is similar to how human language evolves over time through usage, exposure, and cultural shifts. The more people engage with different writing styles, the more they absorb, adapt to, and emulate those patterns in their own expression—whether through comments on social media, reading chapter books, or now, through exposure to generative AI outputs.

​Em Dash Man isn’t here to save AI or destroy it.

He’s just here to say:
          “Maybe we should judge ideas, not dashes.”