4 ms·
For me, getting summaries of meetings is my favorite use case. Saves me from taking notes and I can extract next steps. It also helps me getting started with n
by jagermo 2y ago
For me, getting summaries of meetings is my favorite use case. Saves me from taking notes and I can extract next steps.
It also helps me getting started with new content, kind of building the scaffolding of, say, a blog or social post. It still needs adaption and fine-tuning, but getting rid of a white page is a great help for me.
And I use LLMs to play through ideas and headlines. I would normally do this with other humans, but since working full remote, its a nice sparing partner, although the AI not being able to really give criticism is a bit annoying.
The tools also make it easier to write in English as a non-native, making sure my text does not include any false friends or grammar errors.
- Yizahi 2y agoMeeting summaries is the most hilarious thing these neural networks have produced. I don't know which NN model does Zoom use, but the text it produces is super funny :) . It basically can't parse half of the words, and then generate a random sentences using the remaining ones.
- jagermo 2y agoAgreed, it was super funny, especially if a song played in the beginning or if you switched languages. It has gotten way better, at least in my experience.
- theshackleford 2y ago> although the AI not being able to really give criticism is a bit annoying. I’ve managed to get ChatGPT to a good place in this regard recently and it’s better for it. In fact, it’s been doing such a good job recently that it almost seems like…human like. I’ll have to look at my prompts, but somehow I got it from licking my ass and telling me everything I say and do is amazing to a point now where it almost seems eager to tell me I’m wrong. Which is fantastic, huge improvement. I don’t really use it for coding though, because I am not a programmer. I would have no means today to correctly evaluate 90% of what it would return me.