Hacker News Reader

Comment by XorNot | Hacker News Reader

What is your company? Because honestly: I straight up do not believe this.

Software engineering is 10% knowing how to code and 90% knowing what to code. So I don't buy it either. I use LLMs daily for programming. Sometimes they're helpful. But I see them like one of our first power tools (comparable to the language servers and compilers in our toolbelt).

At BigCos, software engineering is also spending 70% of your day in meetings or handling housekeeping chores in JIRA / Monday / Airtable / Notion / GDocs. I’d love to send an AI agent to meet with PMs to give status updates and take notes if any of their needs change if it could actually do those things efficiently and reliably. Right now, it’d be about the equivalent of having a PM call a phone tree system and leave a voicemail that might get transcribed 85% correctly.

This sounds like you want an AI to turn bullet points into text so that the PM's AI can turn it back into bullet points and vice-versa.

Don't you think that it might be more efficient to just communicate with the bullet points?

Reminds of the Google Translate model that eventually invented its own internal intermediate language. Which would be a nice touch in this kind of applications.

I wouldn't go this far, it's a really hit-or-miss thing, but as someone who is seeing it heavily integrated... Our days are numbered. That number, in my opinion, is closer to a decade, but I have folk at a junior level building out entire internal tools in a couple days what would have taken weeks.

The operations crew is having it automate metrics rapidly. Senior developers are increasing their throughout rates. Things come back to code review in a generally better state because folk say "hey, review this for me".

Is it a real developer? Not at all. Is it affecting our hiring or anything? Nah. But it's already a huge help and rapidly getting better with new models (o3 / deepseek), tooling (CSP and agents), and integration (cline, cursor, memory prompting).

Ten years from now we will see Alexa doing the work of three FTEs output from admin, accounting, data modeling in seconds.

Idea to bank to pull funds to 3d print to market validation in a day.

Not saying obviously so take it from this pizza fan that I've seen some incredible advances.

Example from last week: pasted my terraform that was using internal custom modules and prompt that based on this code build additional services not yet built using the module output so they could all communicate.

First run. 100% accurate. Zero typing after paste and prompt. 2 minutes of time.

It probably saved me an hour at least of looking up module output and documentation with the additional services.

It may be minor or simple yet I now have an additional hour.

Reply on news.ycombinator.com