Consider becoming a paid subscriber to receive my personal notes from building a design-led consumer AI product.

The best thing about making software with AI is also the worst thing — you can make anything. Well, almost anything. Especially if you are working on a consumer creative product like I am, the temptation to go all out is hard to resist. Just this morning, I considered turning the “worlds” in the “inner universe” of Thinkspace into interactive video-game worlds. All I have to do, I thought, is find some open source libraries, character models, environments blablabla, and amp up the experience design. Thankfully, two voices interrupted me.

Read more

Read more

Dear Thinker,

I am here to tell you that using AI does not have to make us spiritually hollow. That human cognition and imagination can and will flourish with machine intelligence, if we start making software with the care of art. That the future of consumer software lies not in the social, but in the deeply personal (so personal that even the human may be unaware of it). That screens should stop hurting, and instead become… magical.

I am passionate about technology for the human inner life, and machine intelligence for the human soul. Before you scoff, hear me out.

Thinkspace and I

My name is Sugandha. Behind me is over a decade long career in product and technical roles. At Netflix, I worked on the product experience of original IPs, at EA I worked on monetising and growing mobile games, at Eightfold I shipped enterprise AI for human talent intelligence. Now, having been through a tiny metamorphosis of my own, I consider myself a “software author” -- a term I am totally making up to describe the new kind of compounded writing x design x engineering craft needed to build technology that matters to us, let alone anyone else.

At some point in my fairly long career, I began to feel disillusioned with technology. I moved to London without the safety net of a job, and started absorbing everything that the creative world here had to offer. It felt like being back to school, as a student of life.

For the last 2 years, I have been pouring my self into a search for truth in the work we do with machines. When I finally found it, I realised truth is personal to each one of us. I am sharing mine today in the form of Thinkspace, a project I am building with more soul and heart than code.

Thinkspace is an Inner Intelligence platform built with the design philosophy of a video game. It turns deep thinking into a spatial, interactive experience, helping you achieve philosophical, creative, and psychological clarity. By giving structure to your inner universe (your “innerverse“), Thinkspace makes “figuring it out” a visual, lived experience grounded in truth. Your truth.

I have imagined, designed, and created every pixel you will see in Thinkspace. I work solo, with Claude as my execution partner. This way, I get to do what I do best or enjoy most (writing, designing, imagining, research, thinking, experiencing, testing, feeling, drawing) and AI does what I don’t do nor enjoy much (write code). One of my hopes is that when you see Thinkspace, you will be compelled to see and use AI in ways that will push the ceiling of your own cognition, expression, and imagination -- areas where machines need us more than we need them.

It has taken me 6 months, 3 pivots, and hundreds (!) of iterations to arrive at the current version of Thinkspace. It has taken me 2 years and many liberating failures to arrive at the current version of myself.

I have come to see how urgently humans need a kind of intervention that can only be described as… spiritual.

Spiritual Intelligence

We live in spiritually painful times. Dark world events and injustices loom on our minds all the time. Arguments and fights with fellow humans batter our souls every day. Fears and uncertainties about the future are a reality of every moment. And with all of that going on, we still have to… grow, improve, be better, plan for the future.

From rings to watches to spas to devices that stick to our foreheads, we are being told to look inwards, not to examine the self the way Socrates described as being necessary for humans, but to measure, monitor, track, analyse, and compare ourselves… in order to improve and optimise the metrics an entire industry is defining for us. This project of self-improvement is not new. Growth mindset gurus taught it. Hustle culture taught it. Therapy-speak taught it. Motivational books preached the hell out of it. Podcasts and speakers sold it in abundance. And for years, we have lived exactly as we were told to live. We purchased the courses, the books, the devices, the supplements. We enrolled in the programs, the cohorts, the fellowships that promised to help us achieve our full potential. We followed the regimens, the rituals, the playbooks, and then some. We dutifully engaged with ghosts of the internet.

Where are we today?

More disconnected with our self, than ever.

More knowledgeable, yet more overwhelmed than ever.

Burdened with gigabytes of data and reports about us, yet unseen.

Overwhelmed and always busy but never quite… there.

Constantly being told by the algorithms who we are, what we should believe, where we should go, how we should live and how we should not.

And of course, “doomscrolling”, “bed rotting”, “quiet quitting”, “binging brainrot”... an entirely new vocabulary of self-description that never existed before our generations. As honest and self-aware, as it is self-deprecatory.

***

Sometimes as a friend and sometimes as an advisor, I have had long conversations with many peers and young people over the last 2 years about their questions, struggles, doubts, and so on. Every time I walked out of a career chat with a younger person, I thought they already knew all that was there to know, but were secretly hoping someone could tell them who they are and what they are meant to do. Whenever I spoke to a friend who did not know what to do next, I realised my job was only to ask the right questions. Every person feeling burnt out needed less talking, more immersing themselves into art, creativity, or some kind of practice that would rekindle their spirit and teach them to live in the present again. Soon enough, there I was, observing all of this in my own journey of self... identification?

As I changed my reading and research direction, I began to see that a lot of us, perhaps all of us, need to read some of the timeless texts across philosophy, spirituality, psychology, literature, neuroscience, and more. Not only because they can help us understand our own brains and often hold the answers to our questions, but to see how far back our questions go in human civilisation at large, and how much treasured knowledge was left by the thinkers and writers who came before us so we could live with clarity, purpose, and meaning.

Somehow, and I don’t know how, all of these experiences, feelings, and beliefs of mine have come together in Thinkspace, to the best of my (and Claude’s) limited abilities. In a way, it is as much a part of my soul, as it is a project of technology and code.

Your Inner Universe… Your Innerverse

You already know this -- an entire universe lives inside each one of us.

For the longest time, computer interfaces had to flatten out our entire selves into profiles, tweets, posts, messages, memes, images, videos. The spatial interface in Thinkspace is a passionate even if imperfect attempt to not only help you see your holistic self clearly and truthfully, but to also reflect your inner universe and its many worlds.

Seeing our deepest inner life is necessary for living in alignment with the self. Such a living -- not in accordance with the world’s norms or an industry’s rules or an economy’s narratives -- but in alignment with our truest inner self is really the only key for humans to live happily, and for the human civilisation to flourish.

Humbly, and this is where some of you would rightfully be skeptical, I believe that machine intelligence can and should aid and heighten human enlightenment and self-knowledge, to enable everyone to live their most clear, fearless, and purposeful life.

3 Core Values

I have designed Thinkspace with 3 core values:

1. Socratic Intelligence, that questions more than it answers and listens more than it speaks. Respects human cognition, deepens clarity.

2. Source and Citation Grounding, in deeply researched frameworks and studies across philosophy, psychology, neurosciences, spirituality, literature, and more. Finds and cites precise knowledge that meets you exactly where your inner life needs it.

3. ✨MAGIC✨ in every interaction with the computer, to nourish human imagination and create a joyful space for inner work that feels therapeutic without being clinical or mechanical.

I have been testing Thinkspace with my own mind and soul for a long time. If I hadn’t experienced its impact on my own creative, soulful, and personal living, truthfully, I wouldn’t be sharing it with you.

I welcome thinkers, writers, readers, gamers, creatives, students and teachers of the human mind and soul to enter Thinkspace with your big inquiries… and a little patience. Your first session is free. Play it whenever you feel ready to sit with your inner voice.

thinkspace.world

With love,
Sugandha

Subscribe now

I will never forget those two unicorns, for example, that both presented with this question in the form of a take-home assignment (in one case) and a live interview (in the other) -- “come up with x number of ways to gamify our app”. A barely disguised idea-hunt, if there ever was one. In both cases, I bowed out of the interview process, not even because I found the practice shady and uninspiring, but because the sweet smell of bullshit was too strong. When I tried to dig a little deeper into what did it even mean to them to “gamify” something that wasn’t close to games, I discovered that they neither understood what made a game a game, nor why they needed to gamify their apps in the first place. I knew what they wanted to hear. I am certain, if I had told them that adding a gacha mechanic was the answer to their monetisation woes, or that adding levels would make their consumers use the app more frequently, they would have been pleased. But I chose common sense. Product Sense 101 dictates that if you are starting with the solution instead of the problem, you are not starting well.



To be inspired by game design is one thing. I’d even argue that everyone in technology should study game design. But “gamification” is as lazy as can be.

Over time, I have come to realise that “gamification” is the tech industry’s oversimplified assumption that because games are highly engaging and retentive, their “formulas” can be applied to any software application to mimic those kinds of user behaviours and metrics. Such assumptions are never backed by any sort of deeper analysis of what makes games tick, what gamers are like as people and as consumers, what sets apart game developers, and how “fun” is written, drawn, engineered, and designed in games. In other words, I have never once met a person who spoke of “gamification” and also understood the depth and breadth of the mechanics, art, and craft that make games. Most people who speak of “gamification” have barely explored games, often because they consider them a waste of time.

So you can guess how pleased I must have been when I discovered and read this most excellent paper by Ian Bogost, aptly titled, “Why Gamification Is Bullshit”.

Ian argues about the emptiness of such “consultant-speak” in words that cannot be bettered. Games are seen by the industry as an “opportunity” whose appeal lies in the possibility of mimicking its success outside of games. And yet, lurks the corporate fear of games as a medium, because it isn’t understood well outside of the gaming community. This combination of appeal and fear has led to the “-ification” suffix that feels, to the average consultant or marketer, presumably safer.

He also dismantles the idea that only games offer real-time feedback to users. The idea that “In a game, anytime you take an action, you receive instant feedback. Positive feedback reinforces good behavior, strategy and tactics, while negative feedback enables you to learn quickly and adjust”, is met with Ian’s solid argument that games are in fact no more interactive or real-time than any other software:

Consider this, the website where you are presently reading my words is also interactive and giving you real-time feedback, either to help you have a better experience or to monetise your behaviour, now or in the future.

There is a particular point Ian makes that I find pertinent to all product engineering: focusing on “gamification” can often distract one from naming the actual thing.

He uses the Toyota Prius as an example, often cited as an example of gamification. The Prius has a dashboard which displays both quantitative and qualitative information about the driver’s style and how it might affect overall fuel economy. This product behaviour has been likened by many to “gamification” because it supposedly makes driving the Prius “feel like a game”. Ian asks if this is indeed the right framing. Instead of loosely analogising this product with games through “gamification”, he suggests it may be more accurate to analyse the Prius dashboard in light of the consumer trend around “business dashboards”.

My own vehement agreement with Bogost’s paper lies in the belief that “gamification” is, more often than not, a weak crutch used when business minds either lack the clarity to name the actual problem and solution, or want to “sexify” something that is otherwise boring. I have been privy to this in my own little tech career. The lust for “gamification” is often deeply personal to managerial folks who wish to appear creative yet strategic to their executives. Ian says it better:

Bogost wrote this in 2014. In 2026, we are witnessing a version of this regurgitate by way of the “taste” discourse. Now, who will write the paper on “Taste Is Bullshit?”

When was the last time you logged into your favorite AI chatbot to talk about a happy day? A promotion you were awarded, a romantic date that went well, a match that led to your favourite team winning, good news you got from your family?

My guess for most people would be never, or well, almost never.

We go to AI to discuss problems, challenges, and issues. If AI indeed has a “companionship” or “friendship” role to play in the modern world, it is only for the messy parts of human life. “Happiness real only when shared” applies to human-to-human relationships. The human-to-AI relationship is as real as it gets, but it is entirely transactional. We need AI when we either don’t have a human we can share things with (loneliness), need answers quickly (convenience), or want to discuss something that requires a non-judgmental space (privacy).


This piece isn’t an analysis of the shoulds and shouldn’ts. There are many philosophical and psychological dimensions to today’s human-AI relationship. I find it best to be honest about which dimension one is and isn’t plugging into. This article does not get into the morality of humans talking to AI. It does, however, get into the psychological drivers that make humans seek AI at all.

I have been researching, discovering, analysing, making, and testing human-AI interfaces this past year and it is obvious that conversational AI is, by far, the only form factor that has really worked in the consumer AI space. Why?

The best way to understand this is to look at where we are using AI most heavily. Below is a list of the top 10 Gen AI uses in 2024 vs 2025.

Consider how much of that aligns with your own use of AI. I find my relationship with AI broadly aligned with this trend, even if I do not subscribe to products that vertically own these use-cases as categories.


Therapy/companionship, meaning-making, organisation of life as the top use cases for AI -- who would have expected this outcome two years ago?

If you zoom out of the AI discourse and look at where society is headed, you might find it aligns with the state of the current world.

It may have started during the pandemic, but over the past few years, the tech consumer became obsessed with self-understanding. So much so that we have developed an entirely new vocabulary of pop psychology (“therapyspeak”), and with that vocabulary, entirely new ways of dissecting and discussing relationships, power dynamics, privilege, and politics of the world we live in. This trend started well before AI became mainstream with the ChatGPT moment in 2022. Case in point: this New Yorker article that talked about the rise of “therapy speak” as far back as in 2021.

So what did AI do in all of this? It arrived at a point when the flames were just about getting started, and fanned them into new social and cultural truths. One of them being the replacement of the “Dr Google” trope with “friend” AI, “companion” AI, “coach” AI, “assistant” AI, “colleague”, “advisor” AI. Second, the rise of “personal AI” and “second brain” AI that lives in people’s apps and chat threads. "


Thus, the consumer who had already developed high literacy in psychological jargon, who prided themselves on the deepthink they sought in writing or reading longform essays and “bangers”, who was starting to adopt “taste” as their new identity, and who loves to think and talk about their thoughts… found in AI exactly what they needed:

• A legitimisation of their obsession with introspection, analysis, and thinking.

• The ability to externalise their head, which they spend most of their time in, making the obsession a lot easier and more indulgent.

***

Anthropic saw it coming, which is why they started pivoting into “thinking” as their brand identity. “Keep Thinking” was the first multi-million-dollar ad campaign Anthropic developed and launched for Claude, in partnership with the independent agency Mother. Here is something very interesting Felix Richter, Global Chief Creative Officer at Mother, said about the campaign:

The film acknowledges our problem-filled present but reframes AI as the solution rather than another threat—and it's true: while the world is full of them, there's never been a better time to have a problem

They were right, too. People rushed to participate in the “thinking” campaign in any way they could -- vying for the merch, queuing up outside cafes, signing up for events in hordes, or just posting about them on the internet.


The young, highly informed, and presumably self-aware consumer who had carefully constructed their entire identity around the workings of their brain finally found an AI brand they could “belong” to, one that validated all the time and energy they spent wrestling with their own head. By legitimising this new-era consumer’s obsession with self, Anthropic managed to create a brand their consumers could connect with. A win-win, if there ever was one. What is consumer culture, after all, if not a quest for creating brands that give people an identity for themselves and their tribe.

It is worth considering that “thinking” is not something we associate with the happy parts of our life. The greatest sources of human happiness are love, health, money, family, nature, spirituality, friendship, and creativity. All of these are “end states” that follow action, not thinking. The human-AI relationship, so far, is not able to offer any of these end-states, because all of them require action and agency on the human’s part. You cannot find love if you do not put yourself out there, you cannot make money if you do not plan your finances or career well, you cannot experience peace if you do not spend time with nature, art, god or your spiritual self. Even creativity, which is the closest AI has come to claiming to offer any of these end-states, remains hotly debated because so far it is the AI that is doing the bulk of creation and instances of it furthering originality of human creativity are yet to be seen. My point being, happiness remains entirely a human responsibility.

What AI is able to assist humans with reasonably well is their self-work -- the first step toward whatever kind of happiness they currently seek. Introspection, planning, consumption, organisation, sense-making… “thinking” is where it begins.


But “thinking” itself is rarely triggered by peace or happiness. Those, in many ways, lie on the other end of thinking. More often than not, “thinking” is triggered by friction. By doubt. By overwhelm. By loneliness. By dissatisfaction. And as long as that remains true, which it likely will for a long time, AI’s role in human life will remain what it already is: a companion for the unresolved parts of life.

For shared happiness, we will always seek human companionship.

The question, then, really isn’t whether AI can lead to human happiness, because humans alone can lead themselves to happiness.

The real questions for anyone developing AI experiences for humans are:

1. How can we make AI extremely good at helping, accompanying, and assisting humans as they navigate the messy or ugly parts of life?
   

2. What new psychological, interactional, and even metaphysical form factors -- other than companionship, friendship, and the like -- can such an AI take?
   

Because “mental health” isn’t the right frame for what is turning out to be a new category of philosophical consumer technology.

Subscribe now

But taste is not a feature. And it is certainly not a strategy.

Why not? That, I suppose, is our inquiry today.


***

If there’s one thing the tech industry loves, it’s a seasonal theory of how the future will work. There’s always a word going around. I’ve lived through a few in my own not-so-long career: agile, lean, gamification, data-driven, personalisation… each arrived like the flavour of the season, only to quietly slip away when the next silver bullet came along.

And now, ’tis the season of taste. Deck the halls, for this season’s miracle cure comes with a sleigh full of vibes.

I am very new to reading Kant, and far from claiming any kind of depth. But the little I understand has felt remarkably relevant to this newfound obsession with taste. I suppose, if we are going to borrow words, we may as well understand them?

Kant (briefly)

Kant believed that taste is:

Subjective yet universal.
When you call something beautiful, you are not simply saying, “I like this.” You are gesturing toward a kind of shared agreement: “Surely this ought to be liked by everyone.” You cannot prove it, but the claim arises anyway. The pleasure is yours; the invitation is collective.

Disinterested.
A judgment of taste, for Kant, cannot be entangled with appetite or utility. If food tastes good because you are hungry, the pleasure is real but not aesthetic -- it is tied to need. Beauty requires a kind of distance, a willingness to appreciate something for its own sake, not for what it does for you.

Purposive without purpose.
His most famous words.
A beautiful object feels as though it were perfectly designed -- yet its beauty serves no function. A rose appears shaped with intention, but nature never designed it for your pleasure. Beauty feels purposive, but is without purpose. It belongs to no agenda.

Now contrast this with the nature of technology, which depends on being:

Manufactured.
Measurable.
Optimised.
Speed-led.
Interest-driven in every pixel.
And above all — scalable.

An industry built on scalability borrowing the language of a sensibility that cannot scale at all. The fixation collapses exactly where it begins.

Subscribe now

Taste × Software

There is a strangeness to taste emerging in the lexicon of technology capitalism. Not because the industry is incapable of sensitivity, but because taste belongs to an economy of attention that software never before had to answer to.

The assertion that “taste is the only moat” misunderstands the architecture of taste. Taste cannot be multiplied across teams, injected into workflows, packaged into frameworks, or standardised into systems. If it becomes operationalised, it ceases to be taste.

And yet, I understand the longing.
In an era where machines can produce infinitely, discernment feels like the last scarce resource. But to imagine taste as a strategic lever is to strip it of the very conditions that make it meaningful.

Taste is neither moat nor mechanism.
It does not scale because its value lies precisely in the fact that it cannot.

That does not make taste irrelevant to software -- only differently placed.

There is power in accepting that not everything of value in technology needs to scale; that some aspects of creation remain anchored to the individual; that sensitivity, judgment, and discernment are not inefficiencies but forms of responsibility.

But the resultant manifestation of taste cannot be limited to cinematic launch videos, templated LUTs and colorgrades, orchestral app graphics, posters and motion graphics, 3D websites and such.

All that is spectacle, the anti-thesis of taste.

What, then?

So, taste cannot scale, cannot be operationalised, cannot be manufactured -- what place can it possibly hold in an industry whose entire existence is built on scalability?

If taste has a role in the future of software, it is not as spectacle -- it is as intention.

It becomes a part of technology that is not meant to scale.

Experiment without agenda.
Express without optimisation.
Make beauty for no reason other than beauty.
Let some choices remain unscalable on purpose.

Most of all, to bring taste, stop trying.
Purposive without purpose.

Taste cannot be the engine of software. But it can steady it. It can guide its hand.
It can remind us -- in an industry obsessed with what is possible -- to also ask what is worth doing.

Her name is Maya Deren.

Born a century ago, Deren remains one of the most original thinkers cinema has ever produced. Not in quantity of work, but in density of insight. Her fingerprints are everywhere in avant-garde film, yet she’s barely discussed in the mainstream. Patriarchy might explain that part better than film history ever will, but never mind.

I didn’t discover Deren early either. But as I have been circling around AI filmmaking -- trying to make sense of the new medium, if it is that -- I keep returning to her. Her perspective of cinema has become the intellectual backbone of how I think about this entire space now.

If you care about art, AI, or the uneasy place between them, read on. This is my interpretation, and so it is bound to be biased, imperfect, and personal, but I think it carries relevance for where we are headed.

Subscribe now

Vertical vs Horizontal: A Film Theory for the AI age

The most counterintuitive insight that Deren brought to filmmaking can perhaps be summed up thusly -- cinema is not just story. She didn’t obsess over plot. She didn’t worship narrative. She didn’t believe cinema was a machine dragging a story forward. She believed film was a way of descending into a moment. A single moment.

She called it vertical filmmaking.

To explain it, she borrowed from poetry and drama. Drama moves horizontally -- one event causing the next, time pushing forward, dramatic intensity climbing until the climax, so on. Time on the X-axis, dramatic intensity on the Y-axis. This, according to Deren, was how popular cinema (Hollywood) worked -- by conforming to a linear narrative structure, where one event leads to another, and another, so on. Think about any film or show you have watched recently, and you might see what she means.

Poetry, on the other hand, moves vertically -- deepening an emotion inside a single moment. And Deren’s view of cinema made it more akin to poetry. In her words:

A poem, to my mind, creates visible or auditory forms for something that is invisible, which is the feeling, or the emotion, or the metaphysical content of a moment. Now it may also include action, but its attack is what I call the ‘vertical’ attack…

Deren’s approach of ‘vertical cinema’ is moment led, rather than being narrative led. Instead of a linear, cause-and-effect story, her approach focuses on ‘vertical investigation’ through techniques like repetition, subjective camera work, and layering, to create a meditative and emotionally rich experience for the viewer. It aims to alter perceptions of time and space to explore an emotional or metaphysical moment from multiple perspectives, rather than just moving a story from beginning to end.

Put simply:
Horizontal is story.
Vertical is meaning.

Horizontal is events.
Vertical is emotion.

Horizontal is where mainstream cinema lives.
Vertical is where avant garde cinema, or Deren’s cinema at least, breathes.

This is where I bring AI into the picture.

Subscribe now

AI Filmmaking Is Breaking Horizontally -- It’s Not a Bug

When I first started experimenting with AI tools, I dragged my ‘horizontal’ brain into every prompt. I wanted scenes to connect. I wanted continuity. I wanted story.

That is where every AI model collapses today.

You get eight seconds. Maybe twenty. Maybe one good frame.

The moment you try to build linear drama with AI, everything starts breaking -- characters morph, angles shift, cameras change, story breaks, styles fail. The medium fights you because you’re forcing it into a grammar it can’t yet hold.

But vertical filmmaking -- Deren’s filmmaking -- doesn’t require continuity. It requires depth and attention. It asks: what would this moment feel like, if you sink into it, instead of pushing it forward?

Vertical cinema doesn’t rely on plot. It relies on feeling.

And that, I am suggesting, might make it weirdly aligned with the limitations of AI video today.

AI Filmmakers are closer to early avant-garde than to Hollywood

A claim I make with some audacity, I am aware.

Who are these ‘AI Filmmakers’, after all?

20-somethings who want to say something and want to be heard but are too impatient and hungry to wait for the industry to give them permission.

People in their middle age who want to replace their monotonous jobs with the work they always wanted to pursue but couldn’t for any number of reasons.

They aren’t trying to be studios. I even think the term ‘AI Filmmakers’ is temporary and ill-suited. They are just people trying to make something that lets them feel alive. They are impatient, hungry, hopeful, not ready to wait for someone to give them a bleak entry point. They want expression, creation, and perhaps, learning. They aren’t anti-industry, it’s just that they are individuals with ideas they are eager to experiment with, using whatever is at their disposal.

Isn’t this what the early avant-garde was? Experimental even if immature? Individual over Industry? Messy, earnest, and human?

If AI filmmaking becomes a movement -- not so much a product category but a cultural shift -- it will come from these people. The ones who will make twenty terrible things before they make one that feels like truth.

Subscribe now

Would Maya Deren have embraced AI?

I suppose there is no way to know, but it may be considered that to Deren and other filmmakers who rebelled through their craft, ‘industry’ perhaps mattered less than art itself. If I may be so bold, I am going to say something that I hope won’t make me feel stupid years later:

The avant-garde filmmakers -- the ones who reimagined cinema in opposition to the mainstream -- might have been kinder to AI than today’s film industry.

Why?

Because they valued intention over production, feeling over polish, experimentation over authority, and the amateur over the institution.

Purely in that sense, maybe AI doesn’t threaten that worldview, instead expands it.

Is it possible that most the resistance we see today comes from those who confuse cinema with industry? As Hollywood once did?

What Should AI Filmmakers Chase?

One thing: originality, not views.

If AI filmmaking becomes another way to generate slop, it will die faster than it arrived. And mind you, slop is not a new phenomenon, the ‘human’ film and TV industry over the past century has produced enough slop to give definition to the word that would have had no meaning otherwise.

But if ‘AI Filmmakers’ use this medium to descend into moments the world has forgotten to see, let alone feel, then -- maybe -- we get new cinema.

A cinema that isn’t ‘horizontal’,
isn’t algorithmic,
isn’t industrial,
isn’t ‘content’,
… a cinema that is ‘vertical’.

I make my pictures for what Hollywood spends on lipstick — Maya Deren

For two decades, we watched software devour industries. Amazon started with selling books online, then turned the internet into a giant commerce flywheel. Facebook upturned the entire advertising industry on its face. Netflix eviscerated Blockbuster and rewired Hollywood. Newspapers became websites, then apps, then blogs and social feeds. The victor’s spoils seemed permanent: whoever owned the code, owned the future.

Then, something strange happened. The most valuable software stopped being software at all.

Subscribe now

Look at what we actually pay for now. Netflix isn’t selling streaming technology, it’s selling Squid Game. Spotify’s algorithm is only table stakes, its playlists and podcasts are the real product. Even Apple, the ultimate hardware company, started producing prestige entertainment with Apple TV+ -- because culture, not circuitry, builds loyalty. Notion, the $500 million productivity darling, markets not databases but the aesthetic of productivity.

When Dollar Shave Club released its $4,500 launch video in March 2012, it wasn’t the razors that went viral -- it was the story. Within 48 hours, 12,000 orders flowed in and crashed the servers. Four years later, Unilever bought the company for $1 billion -- roughly five times its expected annual revenue. What Unilever paid for wasn’t blades or logistics software; it was the audience relationship Dollar Shave Club had built through media.

Or consider Dropbox, whose early animated explainer video helped turn “cloud storage” from a confusing technical concept into something anyone could understand -- before cloud storage became ubiquitous.

Even Facebook, Twitter, and LinkedIn that took the shape of social networking websites became indispensable on the back of user generated content. Yes, content. Social networks became social media, social media became… media.

And now, as we stand at the frontier of AI -- the hottest technology the industry has seen in a while -- the pattern is not only repeating, but spiralling. “Storytelling” is being sold as a service. Narratives are being distributed in essays and reels. Creators, once downstream, are now the lifeblood of the PR economy. Founders are spending more time and dime on producing launch films, posts, and podcasts than on writing software. Content isn’t just promotion, it is infrastructure. The startup ecosystem is flipping the equation faster than ever.

Subscribe now

Foundational models are building slop machines. OpenAI brought AI into everyone’s palms, and then came the Sora app. Grok followed. As did a thousand other AI content models and platforms. Everyone racing for the same destination -- to be the next media platform. One small problem: you can’t be the next media platform until you own the next media.


The Commodification of Software

What happened? Software got commodified. Not all software, of course, but much, much of it did. Once AWS made infrastructure cheap and GitHub made code shareable, and that was pathbreaking. Today, we are in the era of “vibe coding”. Shipping is easier than ever, shaping is harder than ever. It is harder to win a paying customer today than it is to make software. As Andreessen says himself, “the general model for successful tech companies is that they become distribution-centric rather than product-centric.”

And so, “distribution is the only moat”, as the other prophecy goes -- or is it? In a world where every third person is an influencer, where attention is the new currency, what does distribution even mean, if not media?

The Ultimate Inversion

Software did eat the world, but in the process, it may have lost its appetite. The world, as it turns out, runs on something software cannot generate on its own: narrative.

The inversion is nearing completion. Every technology company is trying to become a media company. Big tech already burns billions on content. YouTube fights streaming. Streaming fights games. Games fight apps. VCs fund startups. Startups fund films, essays, newsletters. Creators get paid for media, and then pay media companies for tools to make media.

As AI automates code and commodifies creation, defensibility slips further and further away. The next competitive moat may not even be distribution or data, but story -- owned, embodied, and performed. An industry that ate the world with software is now being eaten by a basic human need: stories and personalities, culture and belonging.

Media may have once meant marketing. Today, media is the product.

Before I jump into this tremendously exciting topic, allow me to start with a quick introduction: I worked in gaming (2017 - 2019) as the product and revenue lead for a portfolio of 10 mobile games in some of EA’s most iconic IPs. I also worked in streaming (2019 - 2021, and 2023 - 2024) where I owned and led launch strategy, product experience, and streaming operations for a Netflix originals portfolio of formats ranging from weekly TV to movies to documentaries to series, and more. Having seen both worlds from close quarters, I am bound to be biased. So, read the below as my very opinionated but passionate take on the future of media and entertainment.

Movies vs Games

As industries, Movies and Games could not be more different.

1. They are consumed differently: Movies are watched, games are played.

2. Formats: Movies are narrative-led, games are interactivity-led.

3. Production: Movies are shot and edited, games programmed and balanced.

4. Distribution and platforms: Movies are discovered in theatres, OTTs, and video platforms. Games are distributed and played on consoles, mobile appstores, and publishing platforms like Steam.

5. But perhaps the single most important difference between these two behemoth industries lies in their audience -- “cinephiles” and “gamers” are separate identities, separate communities, and for all purposes, separate markets.

And yet, there has been a longstanding discourse in the movie industry about games and in the gaming industry about movies: What can one learn from another, how can one tap into the audience of the other, how can one’s IPs be adapted by the other, how can one unlock new formats and genres with the other.

Subscribe now

The discourse has been around for decades, but it has never been more relevant and timely than now. In the present world, both movies and games are being -- for the lack of a less cliché word -- disrupted by AI. From jobs to production pipelines to IPs to talent skillsets to costs, pretty much everything both industries have ever done is being questioned, examined, critiqued or defended (depending on who is doing it).

In theory, a combination of the two should be valuable. Cinema’s narrative strength meeting the interactivity of gaming could theoretically deepen the immersive depth of storytelling. But the actual result is hard to imagine today. We have seen great examples of movie IPs becoming great games, and greater game IPs becoming entertaining movies, but a true marriage of the two as storytelling devices is yet to be seen. The bigger question that no one seems to be asking is this -- what do consumers care for? Even if a new hybrid format does get created, might it land as gimmicky, or would it actually elevate storytelling?

If you are thinking about this, you might like to look into what industry veterans think and say. Some day we’ll look into how filmmakers think about this but for today’s read, I present to you three fascinating quotes from gaming industry veterans. These are clipped from a game design book I hold very dearly (I thrifted it from a rickety shop in Bombay!).

Subscribe now

The first quote is from Noritaka Funamizu, a gaming industry veteran who has designed iconic games like Street Fighter and Resident Evil.

Subscribe now

The second quote is from Peter Molyneux, another industry visionary who has designed games like the Fable series, Populous and more.

The third quote is from Charles Cecil, founder of Revolution Software and creator of one of the greatest point-and-click games, Broken Sword.

And now, I would love to know what you think about this debate.

We live in a hyper-connected, hyper-communicative world. Whether or not we realise it, we are always being spoken to -- by objects, apps, devices, people in our life and in our apps and devices. Imagine the chaos if all of that had to be verbal. Instead of the like button, you’d have to say “I like this” on every internet post. Instead of subscribing, you’d have to tell your favourite creators, “keep sending me everything you ever post”. And how would repost or quote-tweet even work?

I could tell you about the importance of pictograms and their ancientness that dates back to Egyptian hieroglyphics, but you probably already know all that. What I did not know was the pioneering role Japan played in the making of digital pictograms.

The 1964 Tokyo Olympics were a pivotal moment, marking Japan’s post-war recovery and world relations. As they readied to welcome people from all over the world, they had the foresight to know that language would be a barrier. Previous Olympics had used images to depict the various sporting activities. Japan went one step further.

Designers from Japan’s Nippon Design Centre created and debuted the first full set of sports pictograms, that would later go on to become the international standard for all sporting events.

There is a special kind of Japanese pictograms you and I are both very literate in -- “emojis” (e for ‘picture’ + moji for ‘character’). Without emojis, digital communication would not only be boring, but also emotionally flat. Look at the most used emoji in every country. The humble ♥️ is a global phenomenon.

The first emoji set was designed in 1999 for an internet platform that was being developed by Japan’s main mobile carrier, DOCOMO. In case you don’t consider emojis as art, the Museum of Modern Art not so long ago acquired the original collection of 176 emojis for its permanent collection.

Subscribe now

My own tryst with emojis has been both professional and personal.

My first “product” job (circa 2015) was for a messaging mobile app that connected landlords with home-seekers. Chat was the main interface in the product, so we couldn’t afford feature complexity. We needed more people to send the first message because it increased the probability of a conversion into a phone-call (high-intent users). But landlords using the app were older than home-seekers who were mostly young millennials. That’s when I turned to emojis.

A graphic designer and I created a set of “icebreaker emojis” that helped users ask common questions quickly and informally -- “is the flat available?”, “are you a bachelor?”, “is the rent negotiable?” etc. We made them simple but specialised (they had to be different from regular emojis, since users downloaded our app alongside their other messaging apps). We made sure they were fun, age-neutral, and most importantly, easily readable. This was the simplest feature we ever launched and it moved big metrics for the app. More importantly, it taught me that design lives in the perennial tension of fun versus function.

Personally, my favourite emoji has always been the nonchalant shrug. But before 🤷‍♀️ became default behaviour (thanks to Whatsapp), I used to be obsessed with the Japanese “Kaomoji” (kao for ‘face’ + moji for ‘character’) for shrug: ¯\_(ツ)_/¯

In some ways, Kaomojis are even more interesting than emojis. For one, they do so much more with less, turning ASCII into art. Two, they have always managed to be popular yet niche. Even today, every Japanese language keyboard carries Kaomojis, while they thrive largely in anime and manga culture. The nuance and diversity of facial expressions Kaomojis are capable of communicating is a fascinating reminder that good design obscures tremendous complexity.

If the internet is on the brink of a new era, how much space might visual communication get? A recent article in the Japan Times talks about the declining coolness of emojis among young people. Apparently, GenZ and GenAlpha have started finding the full-stop at the end of sentences rude and the once-fun thumbs-up offensive (!). Furthermore, the article says: “while Japanese in their 30s and 40s still sprinkle 絵文字 (emojis) into their messages, the style is now considered おばさん構文 (obasan kōbun, middle-aged women’s style texting).” OUCH.

Pictograms began as empathy, a bridge across languages and countries. Today, they hold the same purpose online, only faster and flatter. They turn words into gestures, tone into pixels. What we design next will decide whether that language brings us together or pushes us apart, or worse, just keeps us scrolling.

Just for fun: try out this super-simple app I “vibe coded” (◔_◔) to play around with pictograms and what makes them:

Pictogram App for Play