Live data from Hacker News

Do your own writing

alexhwoods.com

201–210 of 265 posts

Re: Do your own writing

#201

Earlier quoted context omitted.

Sometimes I don't want creativity though, I'm just not familiar enough with the solution space and I use the LLM as a sort of gradient descent simulator to the right solution to my problem (the LLM which itself used gradient descent when trained, meta, I know). I am not looking for wholly new solutions, just one that fits the problem the best, just as one could Google that information but LLMs save even that searchin…

> I'm just not familiar enough with the solution space Neither is the LLM

If there is something LLMs are good at it's knowing some obscure fact that only 10 other people on this planet know.

Re: Do your own writing

#202

I've long considered writing to be the "last step in thinking". I can't tell you how many times an idea, that was crystal clear in my mind, fell apart the moment I started writing and I realize there were major contradictions I needed to resolve. Likewise I also have numerous times where writing about something loosely and casually revealed to me something that fundamentally changed how I viewed a topic and really co…

I think of it not as "last step in thinking", but as "first contact with reality". Your mind is amazing and lying to you, filling in gaps and telling you everything is ok. The moment you try to export what's in your mind, math stops mathing. So writing is an important exercise.

Re: Do your own writing

#204

Earlier quoted context omitted.

Do you use search engines or do you just memorize all the world’s information?

I don't use search engines for much of anything nowadays (does anybody still?) At work I read documentation if I need to learn something.

This is a very strange and contradictory situation. I'm not sure there's any point in engaging with you since there is nothing but a stream of weak dismissals farming for engagement.

You dismiss LLMs because of factual inaccuracy, which is fair, but now you're doubling down on an anti search engine stance, which is weird, because the modern substitute is letting LLMs either use search engines on your behalf or learn the entire internet with some error and you've dismissed both.

Yes, I'm the "backwards" guy who still uses search engines. We still exist.

Re: Do your own writing

#206
I wonder if anyone from HN would be willing to input their opinion into some features I'm building, which is really outputting LLM generated writing.

It's basically automating release notes and sprint summary's from existing systems like Jira and Linear. The target user is a product team, the target reader are business stakeholders who want to validate your existence. I've found this process to be stupidly time consuming for both our delivery manager, and whichever Dev they decide to tap on the shoulder to help contextualize tickets.

I feel like LLM's are a really good _summarizer_ and it can easily highlight if your tickets don't have enough context for actual people, if even an LLM can't write a summary with good enough context.

Idk, maybe it's a sensible usecase because you REALLY don't want novel ideas from the LLM in this case. You want it to tell you 1:1 what you did this sprint based on a list of issues.

Re: Do your own writing

#209

I wonder if anyone from HN would be willing to input their opinion into some features I'm building, which is really outputting LLM generated writing. It's basically automating release notes and sprint summary's from existing systems like Jira and Linear. The target user is a product team, the target reader are business stakeholders who want to validate your existence. I've found this process to be stupidly time consu…

I'm saying more and more "if you don't have the time to write it then I don't have the time to read it". Therefore my first impression is: if the process is so formulaic that you can automate it, then the content itself cannot be of any interest and the whole song and dance should probably be scraped altogether - think of a person asking ChatGPT "make this one-liner sound professional" and then sending it to someone who auto-summarizes it.

You mention that the target audience is "stakeholders who want to validate your existence", which makes me think that your target audience doesn't really care about what you actually did but rather about being heard. If that's the case then replacing the Delivery Manager (who is arguably doing a good job) with a machine that screams "I want to think about you as little as possible" is definitely a risk. It may work well to provide the DM with a first draft, though.

Disclaimer: I don't know your team nor stakeholders and I'm probably not in your industry.

Re: Do your own writing

#210

I've long considered writing to be the "last step in thinking". I can't tell you how many times an idea, that was crystal clear in my mind, fell apart the moment I started writing and I realize there were major contradictions I needed to resolve. Likewise I also have numerous times where writing about something loosely and casually revealed to me something that fundamentally changed how I viewed a topic and really co…

The Feynman method of solving problems puts a similar emphasis on writing:

1. Write down the problem

2. Think really hard

3. Write down the solution

It's a bit tongue-in-cheek, but there is also truth to it. Step 1 is not optional and actually very important.

Post reply on HN