You must log in or register to comment.
This week, I was talking to someone at work about a topic they know a lot more about. It was a good exchange, I learned a lot. After the meeting, in the span of like half an hour, they sent me ten essay-length LLM outputs.
And it’s just genuinely, like, what is the goal here? It’s cool that they want to provide additional reading material, but clearly they did not proof-read it, so I could’ve just asked an LLM myself, if I deemed that helpful.
There is also just no chance in hell that I’m reading ten essays on the topic, especially when they’re as poorly written as LLM output generally is. I genuinely took a few minutes to try to skim it and I covered maybe 10% of the text, but learned hardly anything.
I just really don’t understand this mindset of “I can generate a lot of text, so I’m being helpful by sending it to you”.
We created an AI usage code of conduct at our workplace. One of the golden rules is “Don’t make other people’s jobs harder.”
As an example, we tried getting AI to generate weekly bug reports from analysing the most common crashes and asserts for the week, assigned to the right person with relevant context. It was a great idea, but too often the bug description and even the title required way too much jargon and comprehension to get through, and often the findings weren’t quite right anyway, so it was only useful as a marker of “look at this crash please”.
These walls of text people share with each other are another great example. If you’re going to do that, you really have to ask it to talk as simply and as briefly as possible with no context assumed. And then read it/correct it yourself!
Exactly this! If I am talking to someone who has expertise in an area, I want to know what they think, not what an LLM outputted.
This sums up my current bugbear with people using Claude, as its output is too verbose. I don’t mind the use of agentic tools, though I want a concise explanation written by a human.
But where it starts to get dicey is when you have both jargon and long text explanations being generated by machines, for people at different levels of understanding. “Bumped dependencies” is good. Everyone knows we are updating packages. “Performed a scheduled dependency refresh as part of ongoing maintenance practices. Minor and patch-level version bumps were applied across the dependency graph, including transitive dependencies where applicable” is bad.
I’ve found success in telling Claude to keep it brief and to use https://en.wikipedia.org/wiki/Simplified_Technical_English
I honestly think this is one of the biggest challenges that needs to be solved. The constant jargon, poor naming choices and assumed prior knowledge of pages of text is extremely fatiguing. It only gets worse as the volume of LLM-generated content in your project increases.
