The Bitter Lessons Behind Kimi Researcher's Taste(medium.com)
medium.com
The Bitter Lessons Behind Kimi Researcher's Taste
https://medium.com/@xinyijin715/maker-story-the-bitter-lessons-behind-kimi-researchers-ui-6654ec66662c
9 comments
I feel this article really needs a proper introduction. It starts "When I first rolled out this Apple-style frontend UI, it looked sleek, and I was genuinely excited." but it never explains what the UI was for. The first picture is of a basketball stats site but it end with "Medications for Menstrual Pain" which is captioned as "interative [?] background". What UI is this post actually talking about?
When you're writing a release, or about some detail of your project, you need to directly give new readers background, or at least provide some navigation to something else that provides background.
A lot of tech and tech-adjacent writing is like that these days. Our cool project X now does Y! With no introduction about what X is, or even a link to something that describes it. So most readers are left to wonder about why should care. At times, I've even tracked down X's github repo only to find that their README has the same problem.
It's an easy trap to fall in when you write. It's easy to feel like you're over-explaining things. You, the author, know what it is, and likely many of the people who follow X know too. The problem comes when some curious person outside of the Xosphere takes a look. Chances are good you'll lose most, if not all, of them.
If you want people to be interested in your X, then make sure they have a way of learning about your X when you introduce improvement Y.
A lot of tech and tech-adjacent writing is like that these days. Our cool project X now does Y! With no introduction about what X is, or even a link to something that describes it. So most readers are left to wonder about why should care. At times, I've even tracked down X's github repo only to find that their README has the same problem.
It's an easy trap to fall in when you write. It's easy to feel like you're over-explaining things. You, the author, know what it is, and likely many of the people who follow X know too. The problem comes when some curious person outside of the Xosphere takes a look. Chances are good you'll lose most, if not all, of them.
If you want people to be interested in your X, then make sure they have a way of learning about your X when you introduce improvement Y.
Yes, if you have never heard of Kimi Research, it is unclear what the article is about. It seems to be referring to Moonshot AI’s product called Kimi.
https://www.moonshot.ai/ https://moonshotai.github.io/Kimi-Researcher/
https://www.moonshot.ai/ https://moonshotai.github.io/Kimi-Researcher/
I am similarly confused. I thought it was about the model, but then it mentions a grid system, typography, JS? Then I was wondering if it was about the editor AWS released, as it was similarly named and maybe I conflated the two, but that doesn't make sense in this context either.
I read the whole thing and still I'm not sure what it's about.
I read the whole thing and still I'm not sure what it's about.
the generated ui that their "deep research" tool uses to present a report
I wonder if this is a problem with the author or the design. I'm not a designer. But I often overthink when programming or writing, and it feels like someone's done the equivalent for design here.
I thought the first picture was pretty good: it seemed like a clean, functional design for a sports game type thing, albeit not hugely original
One way of describing it is "boring", another is "functional, familiar, easy to make sense of, intuitive".
The article reads like teams of people going through millions of tokens and weeks and weeks of manufactured suffering. Maybe it's the writing style, but the time and effort that's gone into "fixing a boring design" - IMO - seems completely disproportional to the scale of the problem, if there even really was one.
Maybe this is why I'm not a designer though.
I thought the first picture was pretty good: it seemed like a clean, functional design for a sports game type thing, albeit not hugely original
One way of describing it is "boring", another is "functional, familiar, easy to make sense of, intuitive".
The article reads like teams of people going through millions of tokens and weeks and weeks of manufactured suffering. Maybe it's the writing style, but the time and effort that's gone into "fixing a boring design" - IMO - seems completely disproportional to the scale of the problem, if there even really was one.
Maybe this is why I'm not a designer though.
This is about the visual design of "Kimi Researcher" (deepresearch equivalent from Kimi Moonshot) rather than the taste of their researchers
Not sure what is it about reads like random text..
I am so bored of this kind of "design". It's just churn for the sake of ego.