San Francisco Experience Overall So far! 20 days In! Part 1

It has been exactly 20 days since I came to San Francisco, and I intend to stay for 40 more. The decision itself was an expensive one to make in the first place. The older I become, the more I think in terms of ROI (return on investment) and taking bets I should be able to navigate. But someone told me that if those are the only kinds of bets I will make, then what is the difference between me and AI? This sparked anger in me, as I wholeheartedly believe that risk and hustle are native to human endeavour.

I have been to the Bay Area before and was truly inspired by the people hustling so hard. The decision was very difficult to justify from an ROI standpoint, given I was paying no rent and living a very easy life back home. So, with no plans and no idea where my career was going, I took the step to be in SF to find other people who are in the same place as me. A little ambitious, a lot hopeful, and ready to take plunges into darkness and come out holding the shining torch that lights the way, it was a VERY BIG STEP, but right now I am so glad I made that choice.

During my flight, I was half-minded to start job hunting the moment I landed. Now, reflecting back after 20 days, that was a dumb idea. I don't know if 2 months is enough time to come up with something and execute it with full rigor. I am going to try my best, and hope that's all it takes.

It is truly inspiring to meet amazing builders, and this journey was no different. I met my cofounder, who urged me to stop thinking of myself as a human being with the potential for emotion, and instead as an MBA who treats everything as a resource. Given that I had done nothing but move from idea to idea in the last 12 months without following through, it felt like surprising advice.

The only thing I could think of before the journey was to build something that was hard to build, a product with deeptech capabilities. But my experience in SF over the last 20 days has been quite different. The takeaway is that any simple consumer startup with a content calendar is far more valuable than a startup building with a "deeptech moat." Deeptech moats aren't really moats in the future. They are going to be part of a level playing field, as intelligent models can understand and augment users' needs much better. The real game, however, is always about a different business model, and how free, uninterrupted intelligence helps people do a lot more than they are used to. Any B2B startup is just waiting for an open-source GitHub repo with a Claude prompt, and fighting it is very hard as more people or agents contribute to building it, making it secure, and maintaining it.

So what is the moat for early stage company? Let me take a step back on someone curious and deeply passionate. The younger me. 

It reminded me to try to become more like the old me rather than the new me. There was a time when I used to write daily blogs on life and tech. What inspired me was not writing or books, it was storytelling, and storytelling has changed its forms. I am not chasing views or followers, I am sharing an idea I am passionate about. The whole perspective shifts when you start thinking from the principles of creating value. You may end up with fewer tech ideas, but your ability to sustain your own motivation through the journey stays much longer. I think storytelling is the biggest moat a startup can have, because the right ideas reiterated across platforms give you the distribution that is the key to unlocking success. If you can be heard and be seen, then 90% of the work is done. The remaining 10% is building a great product.

My next 30 days look like content calendar and community engagements at a deeper level, I want to push for a product that I truly enjoy. 

During my first 2 weeks here, I was working on CSS for Voice, basically a programming language for voice where you can code the voice language for creating voice output. I was able to fine-tune a model called SparkTTS with Qwen architecture where, instead of using emotion tags for entire paragraphs, it applied them to specific words, creating the right emotion and tone for each word. It was a lot of fun, and the results compounded quickly.

The next week, was largely covered with attending hackathons, I built, 

• a chrome extension where you can vibecode any page after it's loaded with custom database, resulting frontend being created as template, which can load on any page. 
• a luma MCP server which would register to LUMA event and also send me details of people who are planning to attend the event, helping me make better choices. 
• a voice agent that would nag me about something 

I found talking to a voice agent about my problems and growth particularly helpful, and decided to move that forward as the project I would prefer to work on full time for a longer duration. I have shown the product to multiple people, created a Google Keep note with a list of signups, and drafted video scripts for the first two videos I intend to publish once the product is in a more polished state.