Thanks for the feedback!
I'll definitely review each one and think if/how can I implement them.
There's definitely some tradeoffs on every decision but I'm on the phase where I'm iterating looking for the best UI/UX.
adrianwaj 23 hours ago [-]
OK. One final idea would be to have a single river-of-items in a readability presentation of full articles. No summaries, just scroll down or jump through each item. Best of luck with it, thanks.
the way you do information density btw is say "include all unique insights, facts, figures, data points, predictions, timelines, and otherwise important information with 100% fidelity"
adrianwaj 1 days ago [-]
Yep, then from that I would have an options panel to toggle only the things you want to see.
The other thing I would do on AIbriefs is have spacebar shift the card-list viewport right to the top and hide the top-most card.
This way reading the site entails loading the page and hitting spacebar after each card is read. This can already be done with Hackerbrain user-script while reading top-comments for any HN item - spacebar or clicking an item's green badge.
Thanks, this is useful. I'm working rn on a toggle to have a dense view.
SiyuanFeng 2 days ago [-]
I love the idea for AI news aggregator website idea. As a data analyst working with AI, it's essential to know the most updated news. And this website helps me catch any AI related news.
michelmi 2 days ago [-]
Thank you! Yes, I'm polling every few minutes and I tried to cover useful sources only.
Please feel free to provide feedback or feature requests.
beeswaxpat 2 days ago [-]
Looks clean and wired up nicely. What model API are you using for the AI summary? Great job!
michelmi 2 days ago [-]
Thanks! I'm using deepseek-flash and gemini-flash. I also tested other ones but these were the best tradeoff in terms of quality and price.
r0fl 2 days ago [-]
This is like a better version of digg
Great execution
michelmi 2 days ago [-]
Thank you! I'm actively working on improving it.
Any feedback or feature requests are welcome.
aleksiy123 2 days ago [-]
The font, the colour, and the dot on the tags all scream Claude slop ui.
matthew_payne 1 days ago [-]
Agreed. Otherwise I like the idea. I’m not an engineer I’m on the sales side of things where I am at, so I am curious and this may be a stupid question but how do you fact check?
michelmi 16 hours ago [-]
I think the misinformation issue here is not as big as it may be in political related news.
Nonetheless, I carefully chose only sources such as official blogs from companies (OpenAI, Google, Anthropic, NVIDIA, etc), tweets from recognized AI figures, or some of the most well known tech newspapers.
There's no specific fact-check feature though.
michelmi 2 days ago [-]
I'm guilty! My background is backend engineer so I relied on claude design for UI decisions. I'll definitely do a review with this in mind.
gneuron 2 days ago [-]
maybe just strip the UI down to basics (techmeme + hacker news are mostly all about the text content)
Have a vertical scrollbar
Add rss
Masonry grid (or offer layout selector) https://share.gemini.google/gAvGtSnQ9i8b
Auto-updates so people can just leave it open in a tab and check without refresh effort
Provide bare-bones brutalist theme option, keeping only vital information
Alternate name: l.A.t.I.v / LaTiV - Vital backwards
Just my tips
And promote sharing through it.
The other thing I would do on AIbriefs is have spacebar shift the card-list viewport right to the top and hide the top-most card.
This way reading the site entails loading the page and hitting spacebar after each card is read. This can already be done with Hackerbrain user-script while reading top-comments for any HN item - spacebar or clicking an item's green badge.
https://greasyfork.org/en/scripts/588721-hackerbrain-sidebar...
Great execution