Hacker News new | ask | show | jobs
by xmprt 24 days ago
Game seems to be heavily weighted towards building side projects. If only you could build a single side project without raising any funds and get acquired for 10M with an 80% cut...
4 comments

Really? I did great just alternating between "grind" and "lay low". Retired with $4.8m at 25!
This seemed to work best.

Unexpected given what I know second-hand about the valley - which is to grind leetcode and keep job hopping.

One of the rarely discussed life skills is going against the flow.
This works only until other people also feel that is the thing to do.
Many strategies have an early adopter advantage.
Sure. And some also have an early adopter disadvantage.
it cannot be discussed or else it becomes the flow.
doesnt that mean your resume will be full of short jobs and you wont get hired anymore
I finished it the first time by just hitting "grind" until burnout got above 80% and then hitting "touch grass" until burnout got down below 30%, repeat until win. It actually feels like my current job.
I tested just "side project", "YC" and "launch", touching grass when burnout is over 80%, ignoring anything work related, trice:

RAT RACE · ACQUIRED sold the company at 25 · walked away with $6.40M · peak burnout 82%

RAT RACE · ACQUIRED sold the company at 26 · walked away with $11.37M · peak burnout 91%

RAT RACE · ACQUIRED sold the company at 26 · walked away with $8.96M · peak burnout 95%

Trying to balance grind, touch grass and side project took me 20 years before retiring with 20'ish millions after selling a hard-worked side project.

It seems that if you ignore work and do "side project" only from first month, your networth increases with the compensation when you are fired, and if you last one year (almost all runs), you can keep "side project" indefinitely.

Yes, I really don't think that such a high fraction of people get purchased for fucking 15M after one year lol
Well thats your problem, everyone knows you got to apply to YC to get funded

/s

this is possible now with AI
In all seriousness, where are these stories? There are so many "entrepreneurs" vibing away, but who is getting paid? I feel like there should have been some decent exits by now but all I've found are a few lifestyle businesses and a lot of hucksters.

My experience: I feel like I could build a real product solo much more quickly than ever before, but the reality is my side-projects have been mostly futzing around with coding agents and related infrastructure - like building my own command proxy system with LLM review/classification & deterministic approval policy - for sandboxed agents to manage cloud infrastructure in controlled ways - building the builder who never builds. I see a lot of that kind of stuff on Twitter too.

from a business standpoint, PMF was never about the underlying software. It’s a developer’s wet dream to toil away in a garage toward some Technical ideal, and then have the world applaud their genius and shower them with money.

Now to be fair, there are windows of time where that really did happen for some few. As a craftsman developer, I have the same wet dream. The problem is I know it’s just a dream now that I’m older.

So what AI has done is condense the frame in which a developer can spend years toiling away in the belief that they just need to keep going. Now the feedback loop is more instantly connected to reality and the reality is most all this stuff nobody wants or needs. starting a business now is ironically about all the business bits and that’s just rather annoying for builders like the HN crowd, myself included. This is more cathartic than anything. What do I know about entrepreneurial success?

Something that I was hoping was that LLMs would make software development easy enough so that I could upskill in areas that determine entrepreneurial success. But I find that the quality of code from LLMs is low enough that I have to spend a lot of time doing QA (or manual coding).

But also you could write the god damn best software out there but if you can't convince people to open the web page or download the app then it goes nowhere. Absolutely nowhere.

Because code was never the bottleneck.