recently I get the feeling that the entrepreneurship ecosystem is once again falling into the divide of the pre-covid era: does it really matter where you’re building from?
\n\nthere’s definitely a “big bay migration” underway - technical builders find themselves in either SF Bay Area, or the Chinese Greater Bay Area. Those who make the move swear by it.
\n\na flower is ever so often given as the example on the other camp: even in the harshest of environment, be it a sidewalk or a dam wall, the resilient one will find a way to grow
\n\nmy humble take is that we are simply at a point where, once again, we refuse to accept two truths at the same time.
\n\nI think the difference manifests itself in the first dive and the last overarm.
\n\n1/ the dive
\n\nif you’re doing something that has the product zeitgeist fit from the get go, you will go for it anywhere.
\n\nif it has precedent, you can only build on it. Truth is, you don’t need to be at the breaking edge of innovation to build a B2B marketing SaaS.
\n\nif you have decades of experience in the swim lane you’re diving into, you don’t need to be with the crazy builders.
\n\nalas, if you’re a daring builder, this all changes. If you’re diving into the unknown, you will benefit from a few more blind swimmers by your side - if for nothing, then for the simple fact that courage is contagious.
\n\nif you feel that you’re way out of your depth in your venture, you will benefit from a few more venturers in doubt of their reach - if for nothing, then for the encouragement you will get from seeing, loud and clear, that their heads are still well above water, and so is yours.
\n\n2/ the last overarm
\n\nthe dive plunges you in. the overarm keeps you from sinking - scaling, selling, grinding through the churn until you reach the coast.
\n\nflower grit cracks concrete. fine for survival. but plants? they hunger for soil that feeds the stretch - networks dense enough to share oxygen, VCs who nod at your weird, talent that vibes your stroke.
\n\nonce again, if you’re in the first category of builders, row solo from anywhere, APIs and freelancers are borderless. but if your bet’s at the bleeding edge, thin air starves you.
\n\ntruth is, whenever I venture into something entirely unknown, it is the last stretch that pushes me the most. even when the product is built, the user surveys are conducted, and the brand is ready; the feeling of it all being utter nonsense creeps in.
\n\nit is at those moments that I need a soil that not only allows for survival, but also pushes me to bloom. that comes in the shape of others who stay up late on Sunday nights with you in a random WeWork just like you, bartenders who bring you a beer when you have to whip out the laptop at 2 am in a bar, and friends who make it easier for luck to find you and harder for falls to break you.
\n\nat the end of the day,\ni don’t think you need to upend your entire life in pursuit of greener grass - but that does not mean you can’t reach over every now and then to take it in.
\n\nthat, at least, is what I intend to do in these next few months, as I quit my job, and set sail for a trip to the promised lands in which the flowers bloom the brightest - just to see what they do right, and hopefully bring some of it back with me.
\n", "date_published": "2025-12-28T00:00:00+01:00" }, { "id": "https://www.karacaufuk.com/thoughts/building-nonsense-products/", "url": "https://www.karacaufuk.com/thoughts/building-nonsense-products/", "title": "on building nonsense products", "content_html": "LLMs and foundation models are all the rage right now, and understandably so. They are powerful tools that can be used to build a wide variety of applications, and they have the potential to revolutionize many industries.
\n\nAs is the case with most frontier technologies, the ones building the strongest products on top of it are those who are building the entire vertical. With the AI boom, this vertical requires an insane amount of resources to build and maintain, which is why we see so many startups trying to build quick wrappers around these models to capture some of the market share.
\n\nThey are easy to dismiss as “nonsense products”, but I think they serve a purpose in the grand scheme of things - for the innovative bunch. They are built to grab attention and capital, which can then be funneled into more meaningful and innovative projects.
\n\nI’d expect, past this initial stretch, foundation models to be a not-so-innovative commodity that is taken as a given the same way we treat hosting, or cloud compute right now.
\n\nSure, current major AI companies provide both the platform and the product, but when (and if) we reach the development plateau for the underlying platform, then it will be those providing the correct tweaks/incremental improvements who will prevail as customer favorites.
\n\nI agree with those online that most of the wrappers striving for revenue-optimization over sustained PMF will be gone one way or the other as the base technology settles.
\n\nIn my humble opinion the smart founder of these companies will do one of the two things till then:
\n\nThe odd founder (exceptional or the exact opposite) will hold onto the product and potentially make something cool out of it.
\n\nEver the example: Instagram and TikTok didn’t necessarily kill Snapchat, but they sure did not help its case either. Regardless, it managed to survive and thrive in its own niche.
\n\nPersonally,\nI’m a big fan of building companies that matter and not clinging onto most products too tightly.
\n\nIf the means to starting that company off is to build a few not-so-groundbreaking products that payroll actual innovation later on, I’m all for it.
\n\nThey are the companies I believe can create the right conditions for good progress. They’re the ones I’d bet on.
\n\nFor all my rationale, there will always be exceptional cases, fueled by strong convictions about what a product can be, and driven by the founders’ desire to prove those calling it “nonsense” wrong.
\n\nThose are the ones that create great progress. Those are the ones I root for.
\n", "date_published": "2025-09-27T00:00:00+02:00" }, { "id": "https://www.karacaufuk.com/thoughts/knowledge-and-communication/", "url": "https://www.karacaufuk.com/thoughts/knowledge-and-communication/", "title": "knowledge and communication", "content_html": "This is uploaded from a late-night note in early 2023 from before this website in its current form existed - but it feels relevant to share now (late 2025). My convictions in this text still hold true, however, I see the fault in thinking that we can ever fully capture and transfer knowledge without loss, for loss in itself is knowledge.
\n\nKnowledge is an interesting concept. Ever since the earliest of days, our species have relied on some form of knowledge transfer to continue to their existence.
\n\nIn the beginning, it was oral traditions, folklore, and stories of kinds. Over the years and through the people, these stories and traditions have evolved or got lost. Those that evolved were the ones we were able to make sense of within our contexts and environments at the time. The ones that got lost were either not relevant anymore or were too complex to be passed on effectively.
\n\nThrough progress, we moved onto the written form. It preserved the words, but not necessarily the context or the knowledge. In non-didactic forms, the knowledge transfer was still lossy, for what was written rarely ever the full picture of the knowledge that was to be transferred. The written form did help in preserving knowledge better than oral traditions, but it was still not perfect.
\n\nOver time, we developed more structured ways of transferring knowledge - books, articles, papers, and later on, digital forms like blogs, videos, and podcasts. Alas, even with these advancements, we have simply enriched what I like to call the “created” form.
\n\nThe “created” form is the one that is created for the purpose of transferring knowledge. It is often a simplified version of the original knowledge, tailored to the audience’s understanding and context. The biggest pitfall that it introduces is the preconditioning that the author itself is not aware of.
\n\nIt wouldn’t be too large a claim to say that we are simply not capable of accessing the full breadth of knowledge contained within our information processing systems (brains) to be able to transfer it in its entirety. The links within blobs of information are often lost as we try and decipher and transfer them from one brain to another.
\n\nOur rich, audio-visually enhanced documents, in my opinion, are nothing but a larger context means of transferring the same “created” form of knowledge. They are still lossy, and they still suffer from the same preconditioning problem wherein the author is not aware of all the links and contexts that they themselves have.
\n\nNow, as we reach the age of AI, I am slightly disappointed to see us spending the vast resources at our disposal in making the transfer and management of the “created” form more efficient. It is a worthy pursuit, for it enhances our ability to share and access knowledge - but it is nowhere near the potential that we could be reaching for.
\n\nI do find that our current means of knowledge transfer is the single biggest bottleneck in our progress as a species.
\n\nYet they are all limited in their bandwidth of device-to-brain. The “created” form is still the main medium of transfer.
\n\nI believe that the next frontier is to be able to transfer knowledge in its more raw form. The form that is closer to the original knowledge, with less loss and preconditioning. The form that is more aligned with the way our brains process and store information, without the need for translation into a “created” form.
\n\nI don’t precisely know what that form is yet, but I believe it is out there. It could be a new technology, a new medium, or a new way of thinking. It could be something that we haven’t even imagined yet.
\n\nI envision it as a standardized knowledge format. A format, quantum in its essence, that need not be known in its contents or entirety but rather in its existence and structure for it to be useful. A format that can be easily transferred, stored, and accessed by any brain, without the need for translation or interpretation - neither to a language nor a medium. A format that forms links in any brain that accesses it, without the need for the author to be aware of them.
\n\nThose who achieve this will be the ones who will unlock the next level of progress for our species. I’d chronologically expect them to:
\nAny step in this direction will be one I will drop everything to be a part of.
\n\nUntil then, I will continue to consume and create the “created” form of knowledge, while dreaming of the day when we can transcend it.
\n", "date_published": "2023-05-12T00:00:00+02:00" } ] }