It's 10.10.2025 saturday morning and It's been 6 months I have been building my first startup, www.usedesktop.com, A computer use agent that works on your PC from anywhere.

The product is barely working and there are still a lot of features I need to implement. But that's not the biggest problem I have right now. The biggest challenge I face is that the core technology needed to make computer use agent so good that people are willing to pay, isn't there yet.

I mean if you think about Anthropic's CUA still in beta even after its initial launch in 11.2024 and the fact that other frontier labs ard struggling as well, we can already assume that we need more time until it gets more reliable.

(well, at least gemini launched their CUA API 2 days ago, so it seems like a lot of big techs are keeping eyeing on CUA at least)

thus, at this moment, I am essentially building a business for a market that doesn't exist(yet?)

Why isn't there market? 

Nobody at this moment will pay money to use my service or CUA in general. Even I wouldn't because all the CUAs are slow, clunky, not accurate due to LLM's non-deterministic nature, expensive and overall unreliable.

Some people I know of told me if you are building for a market that hasn't been discovered, it probably means because it's a dead one or not worthwhile pursing it. (This also reminds me of YC cofounder Paul graham criticizing founders doing "Solution in search of Problem)

And I mean it's not wrong, If you are building something that isn't sellable since there is no market then it's more like a sideproject than running a company.

However, I am still building it because I believe in the vision. 

CUA is fundamentally designed to replace human workers. It should basically do any tasks on computers that human could do. Some people might argue on this whether it is for assisting humans or replacing humans but its ultimate goal is surely to replace human.

And since GPT outperformed the smartest people in the world I believe it's just matter of time when the technology needed arrives.

In Peter Thiel's Zero to One, there are some famous quotes I like

- Competition is for losers

- Each moment happens once

If I build something that already has a huge market and proven, then I am joining a competition hell. 

Since the market is verified, a lot of small and big competitiors will exist already and join. 

Would I be able to sell my service? Yes, probably but would I survive? Well, probably not.

Furthermore, I generally like to do stuffs that I like of. I honestly don't want to build stuff for others which I don't enjoy making. 

I believe there is a vision but the market isn't ready for me to sell to it yet due to the technology not being there yet.

It's very lonely journey but fortunately one of my great strength is consistency.

When I say, I will do something. I always do it. When I wanted to hit master in league from Gold 4, I did it in 1 year. When I wanted to study topology, I did it everyday for 160+ days without a single break even when I was sick, on weekend and on my birthday and posted what I learnt on reddit https://www.reddit.com/r/TIL_Math/ 

Antrhopic, Google, Meta, Microsoft, OpenAI and some big names have all officially announced that they are joining this market. Compared to them, I am just a solo entrpreneur but this doesn't make me nervous or doubt myself. I have always done what I said I will do and I am pretty sure I can beat them.