Viewing profile — fbrncci
fbrncci
HN member- Joined
- Mon, Sep 30, 2019, 3:37 AM UTC
- HN karma
- 521
- Public activity
- 224 items
- HN profile
- View on Hacker News ↗
About fbrncci
Recent public activity
-
comment
Comment #49162374
At this stage with AI and how massive AI generated codebases are growing, the chance of fixing anything manually has already reached 0.
-
comment
Comment #49108382
“The bubble ending” is possible the biggest meme of 2026 (and onwards).
-
comment
Comment #49027947
My main market is still US. Largely US founders with startups and companies.
-
comment
Comment #49027685
I am a small entity, expecting less than $250k in yearly revenue. Somehow I’d be honored to be on that list. I wouldn’t expect to end up there either. And I could just accept crypt…
-
comment
Comment #49027498
Living outside of the US (and EU). I’m in the process of launching an agent assisted automated pentesting and authorized offensive security service. I’m using open weights models t…
-
comment
Comment #48986322
Sure if you want to bake a cake you also need to invent the universe first… As long open weight models get released; no moats are going to exist. Both windows and Linux are weak co…
-
comment
Comment #48985691
Due to all their fumbling I decided to not even try Fable. Why get used to a model with limited shelve life when I can just use the older models that do well for me? Well then 5.6 …
-
comment
Comment #48985641
Safe to say there is no moat.
-
comment
Comment #48985636
Azure should have those.
-
comment
Comment #48865927
Instead the same workers would be unemployed if they stopped doing these kind of jobs all together. There are few options in the Philippines, and BPO is actually a well respected c…
- comment
-
comment
Comment #48839765
Don’t be naive. Every model provider does this.
-
comment
Comment #48812332
Have you tried accepting less than 200k? Sub 100k maybe. Works for me (at 40+). And then just work several clients as contractor. Beats flipping burgers, comes out at more than 200…
-
comment
Comment #48811934
AI still can’t fake a 50$ background check and verifiable references from past employers or clients you worked for. So I wouldn’t worry about juniors masquerading as seniors. Doing…
-
comment
Comment #48789244
That does make it sound like there are only 2 types of software. The “serious” and the “unserious”. There’s probably a few dozen types rather, it’s much less black and white. At a …
-
comment
Comment #48769571
Sounds like a ragequit to me.
-
comment
Comment #48713337
Anyone that has played civ long enough has done something like that too. I just start pillaging farm land as well, but of course a nuclear strike when you can no longer win would a…
-
comment
Comment #48680757
The tokens/salary ratio is not relevant at all. Because while 200-500$ is a lot of money, it’s still a fraction of the salary you’d pay any dev in the world. It just comes out as a…
-
comment
Comment #48667675
My costs are 0$ as any token or subscription spend on agents is invoiced as an expense to my clients.
-
comment
Comment #48667279
You made me realize something. I routinely spend upwards of 500$ per month on LLMs for coding (expensed towards clients). However I live in a place where 500$ is around the avg. sa…
-
comment
Comment #48601582
Are you actually seeing any signs that we’re going back to how software was written before, and needing those skills in the same way? Because I sure am not seeing that right now. A…
- comment
-
comment
Comment #48563059
I am still trying to find a replacement for cursor. But in the past 30D my automations and my own coding has consumed well above 10 billion tokens for less than 300$ with auto and …
-
comment
Comment #48430775
Well to a lot of people this is still a foreign concept.
-
comment
Comment #48103492
Why Python? Because I have written it for 10+ years, know how to debug it and I can smell it within 10 seconds of the agent writing code if it does something that is going to end i…