- It's an ever evolving information repository - the initial use - from Wikipedia to blogs to newspapers.
- It's a debate space - forums ( used to be newsgroups )
- It's a transaction space - ecommerce, marketplaces
- It's a social space – from keeping in touch to meeting new people – social media, dating websites. used to be irc
- It's an entertainment space - tiktok, youtube, netflix, etc...
AI will have the harshest initial impact on the information repository use. It will cannibalize it but also needs it to feed itself.
The transaction space will be affected. Protocols like MCPs once strengthened will need to support transactions. Payment infrastructure will need to be built for this.
Then, the social space will be the weirdest. AI Companions will become ubiquitous, naturally filling the void left by the weakening of the social fabric and loneliness epidemic.
For the debate space, 99% of it doesn't play the role of debate, but more of the role of echo chamber and social validation. It's AI Companionship but by community. These spaces will stay. AI is one to one, not one to many. But they will drastically lose appeal. AI will perfectly play this role of validation and echo chamber.
Finally, entertainment is already being disrupted. The question will be how the industry as a whole ( it's more than purely content creation, it's the whole mythos creation around it ) will adapt to the possibility of on the fly content creation.
AI will become the main human-machine interface, and the role of machines will grow exponentially in our daily lives. The capitalistic concentration that will ensue will be never seen before. The company who will win AI will be the most powerful company in history. They will dominate not only tech, but culture, economics, world view.
Remember, GPT2 was only released 6 years ago.
Identity is a seductive illusion. It promises meaning, coherence, and comfort. Yet beneath this comfort lies a quiet tyranny. Once we define ourselves, we imprison ourselves. The fixed “I” we cling to doesn’t stand scrutiny, not scientifically, not philosophically, not psychologically.
Consider neuroscience. The self, the stable “I,” is nowhere to be found in the brain. What scientists find instead are fleeting processes, overlapping patterns, temporary states. Neuroscientist Anil Seth, in his book “Being You,” argues convincingly that the self is a kind of ongoing hallucination—a comforting fiction the mind generates moment by moment. There’s no stable core beneath it, only fragments, constantly rearranging.
Philosophy has long understood this. David Hume, centuries ago, searched for the self within consciousness and found only sensations, perceptions, shifting sands. Nietzsche echoed this idea, mocking the notion of a stable self as merely a trick of language, a grammar-born illusion. Ludwig Wittgenstein, too, dismantled the belief in private inner truths. For him, meaning emerges not from introspection, but from the shared dance of language and action. To think there’s a stable self hidden deep inside is to misunderstand how language, thought, and reality interact.
Freud, despite the distortions of modern therapy, never saw us as coherent beings either. For him, the human psyche was a theater of conflict, desires clashing beneath consciousness, forever hidden from full awareness. Freud’s contribution wasn’t mechanizing the mind, but revealing its profound contradictions. Unlike later theorists such as Lacan, who mistakenly reduced humans to mere patterns of meaning—Freud knew that beneath our carefully narrated lives lay irrationality and mystery, impossible to tame fully by analysis.
Yet today’s therapeutic-industrial complex tries precisely that. Therapy, once an exploration of human depth, now resembles a dating app business model. Platforms like BetterHelp and Talkspace profit not by curing distress, but by prolonging it. You pay, endlessly, to decode yourself. Success for them isn’t resolution, but endless self-inventory, endless labeling.
This is where the concept of “mental health” itself becomes subtly oppressive. It implies an ideal state, a normal range, a standard we should measure ourselves against. Ian Hacking described this phenomenon as the “looping effect”: labels shape behaviors, behaviors reinforce labels, and individuals get trapped in a feedback loop of self-pathology. When identity becomes a medical diagnosis, we surrender our autonomy, our ambiguity, and ultimately our freedom.
Trauma, too, has morphed into social currency. It’s become the dominant metaphor for interpreting every aspect of life: attraction, friendships, career choices. James Davies, author of “Cracked,” highlights how the psychiatric model strips our lives of meaning, reducing complexity to clinical symptoms. But not every strong feeling is a wound; not every quirk a disorder. Of course, painful things do happen—sad things, terrible things—but we speak as if these things always happen to us, placing ourselves at the center of a chain of causality and consequence. In reality, events simply happen. There is no inherent meaning, no cosmic intention. How we weave these accidents into our personal story is entirely our decision. How we let them shape or influence us is our choice alone. It is precisely here, in this decision, that our ultimate freedom lies.
Here’s the core of the matter, the radical truth that’s rarely stated plainly: there is no fixed identity to discover, no true self hidden deep inside. Identity is a myth we invent because life feels safer if we pretend there’s something stable beneath the chaos. But life resists stability. We are never fully known—not even to ourselves. The moment we accept this, the instant we let go of trying to pin ourselves down, we become free to live.
Happiness is not the reward of endless self-exploration. It’s not a puzzle we solve or a mystery we decode. Happiness is a decision, a leap, an act of rebellion against the tyranny of labels, against the logic of suffering. We don’t find happiness by understanding every trauma or categorizing every impulse, but by letting go, by accepting uncertainty, by choosing desire, will, and joy despite everything. This is what Kierkegaard called the leap of faith—not into religious belief necessarily, but into life itself.
Ultimately, identity is not something we have. It is something we do, something we perform, something we constantly recreate. The human condition is not static being but perpetual becoming. Nietzsche said it best: we must become who we are. And in doing so, we reject the cage of identity entirely. We are not diagnoses, symptoms, or personalities. We are simply alive, driven by desire, moved by joy, ever-changing and never fully defined.
Yet this does not mean giving up on growth or self-improvement—quite the opposite. It is precisely because we are not limited by a static identity that we can truly act, improve, evolve. But nobody can do this for us. Nobody can save us from ourselves, not a therapist, not a romantic partner. The responsibility for our existence and happiness lies entirely in our hands.
By letting go of identity, we open ourselves up to genuine change. Free from the suffocating need for recognition, our relationships become richer and more authentic. Most relationships are neurotic precisely because they are driven by demands for validation rather than true acceptance. When we stop begging for recognition, when we let go of the myth of identity, we allow ourselves, and others, to exist in all our beautiful complexity. Relationships cease to be transactions of self-worth and become genuine encounters: open, curious, generous.
In the end, perhaps the bravest act is not endlessly “working on ourselves,” but simply living—boldly, messily, without apology or explanation.
Just made a test for our use case and the results are pretty good! I have an ecommerce company selling fine food and one major issue is training our employees with deep fine food knowledge.
Coupled with a RAG / Internal documentation, it could generate training material s for also internal procedures, etc...
One thing I noticed, is there is a lot of repetition of the same concepts in the slides generated. It would be great to be able to tweak the outline before it generates the slides. But all in all, really great stuff! Congratulations!
All the incredible performance and success stories always come from these Twitter posts, I do find value in asking simple but tedious task like a small refactor or generate commands, but this "AI takes the wheel level" does not feel real.
The way I prompt it is first I write the documentation of the module I want, following the format I detailed inbthe master documents, and ask him to follow the documentation and specs.
I use cursor as well, but more as an assistant when I work on the architecture pieces.
But I would never let an AI the driver seat for building the architecture and making tech decisions.
Backend: Elixir & Phoenix
Database: PostgreSQL with Supabase (originally CockroachDB, big mistake)
Auth: Supabase
Frontend: Next.js
UI Components: shadcn
GraphQL API: Absinthe
Hosting: Fly.io (Phoenix) + Vercel (Next.js)
Storage: Cloudflare R2 + CDNLittle note: It seems the search is only by book title, not by author and not resilient to typos.
- Images are super small.
- You can do exactly the same level of performance with http://instantclick.io/ to prefetch pages and aggressively cache content on the backend.
- The only dynamic functions are the cart ( session ) and the search. The rest is just navigation.
- There is quite no content
If we compare to the original:
This leaves out what is imho the most impressive part of McMaster website: the deep taxonomy of products and super detailed search with custom criteria per product type and sub type. This is the part that is the most amazing for me and the most complex on an e-commerce website to build *AND MAINTAIN* over time.
If we compare to normal e-commerce use-case, it's lacking a lot of features that have deep impact on website speed:
- No analytics and marketing tracking ( ecommerce without tracking is not realistic performance wise )
- No image gallery, no high resolution images
- No product description
- No product recommendation
- No faceted search
etc...
Craft is great but a bit slow in my tests. They have a really powerful e-commerce offer though and is really flexible.