Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
StahlGuo
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
3 ms
·
1.
▲
by
StahlGuo
3mo ago
Recently, my experience is that the hardest part of writing Enterprise document by AI is not how to generate a word or excel, but to generate a office document that is accountable. First draft generation is just a small part of the whole w
2.
▲
by
StahlGuo
3mo ago
I agree to split the "Brain" and "Hand" of an agent,many deficits of agent workflows is not from what LLM don't know what to do, but from LLM getting too much authority to edit the world it belongs to and break the
3.
▲
by
StahlGuo
3mo ago
Beside from particular project, I would like to talk about my rule of thumb in agent usage, which is to separate “smart” from “authorized.” That is, to build a state wrapper and separate agent from making runtime decision, a dangerous desig
4.
▲
by
StahlGuo
3mo ago
You are right, my solution is to implement real world software engineering process, something closer to a real Software engineering team.I split multi agents into one repo and gives them different tasks"requirements check, product/
5.
▲
by
StahlGuo
3mo ago
that sounds like an interesting idea.
6.
▲
by
StahlGuo
3mo ago
Broswer contains many security concerns, how do you solve privacy and security.
7.
▲
by
StahlGuo
3mo ago
claude code + codex,for sure. I also actively try Other runtimes.
8.
▲
by
StahlGuo
3mo ago
what a cute project
9.
▲
by
StahlGuo
4mo ago
I would said electricity generation, Solar+Energy Storage will be the solution of human future in the coming 20 years.
10.
▲
by
StahlGuo
4mo ago
Wow, is this like a social media for Vibe coder?
11.
▲
by
StahlGuo
4mo ago
I would try it today, sounds good
12.
▲
by
StahlGuo
4mo ago
Wow,that sounds gread for the sport Thanks Fable, we miss you
13.
▲
by
StahlGuo
4mo ago
I don't mean never check it by myself, but I want to discuss about a methodology to ensure AI generated text has the same auditability and tracebility like human wrote text.
14.
▲
by
StahlGuo
4mo ago
Yes, i agree that quantitative part, like dates, numbers, amounts should be extracted and let the LLM to output original numbers and computation steps. That's not hard for a briefing harness. However, my most confusion part is qualita
15.
▲
by
StahlGuo
4mo ago
Oh no I am not a attorney, but I need to analyze the market, 20-F and 10-K recurringly and make strategy suggestions, and so yes I use LLM to do business brief.
16.
▲
by
StahlGuo
4mo ago
I am sorry that I did not explain my question explicitly. I write recurring weekly briefings for internal use, such as market insights and industry news. I’m not trying to make AI-generated text “believable”. I’m asking almost the opposite
17.
▲
Ask HN: How do you make LLM generated text believable?
3 points
by
StahlGuo
4mo ago
|
9 comments