Stop saying formerly known as Twitter. We know what X is.
2026-06-10 04:35:53
10
(not) dave :
Why is no one talking about TikTok's parent company having an open source recommendation system too, I feel like it's important topic
2026-06-15 11:26:27
1
ntncy :
It'd be awesome if users had the freedom of choosing the optimization objectives for their platform experience (eg, "optimize my feed for" engagement or truth or balance or serendipity). The closest I've read someone getting to this was Bluesky, supporting custom/novel algorithms to be imported by users as plugins.
2026-06-10 11:16:10
7
professor_light :
How could an algorithm know truth?
2026-06-10 18:53:55
6
justin :
I already know it. If (elon tweet) then insert_in_slot_1
2026-06-10 12:48:50
20
forlaughs :
also This vid is cool but I lowk don’t understand anything 😁 sm tech jargon im lost unfortunately
2026-06-09 23:41:07
2
Gunnawunna768 :
This is old news? The repo has been public for a while??
2026-06-10 03:57:41
18
Shelly Belly :
An algo can’t determine the truth, it can only determine what others say is the truth. What a dangerous route that takes.
2026-06-14 20:48:15
1
Danny Diaz :
Thanks for this update.
2026-06-09 23:06:31
10
Suhail Khalid226 :
Thanks for the insight and engaging story telling
2026-06-13 06:34:45
1
Gresto 🏳️🌈🇺🇲 in 🇬🇧 :
And is there a reason we believe this is the version X is actually using?
2026-06-10 12:07:20
5
Usr98681 :
Not tech minded, but would find it interesting to know how such algorithms weight clicks. Like watching 10secs vs a block or dislike of rage bait?
2026-06-10 06:52:31
1
QingShiro :
what does this now mean for users?
2026-06-10 11:58:28
0
Fregmnt🇰🇵 :
Twitter has gotten so bad
2026-06-10 14:57:59
1
Fardarterfu :
Where is the insertion of Elon Musk’s tweets into my feed?
2026-06-10 05:03:12
4
djdkgkenclrnckfk :
What's the definition of quality?
2026-06-10 05:24:09
3
Cheesehead :
You look gorgeous.
2026-06-10 01:33:08
5
Jae Williams :
As a software engineer this is a really good video. I feel even a non tech person could follow this easily.
2026-06-10 16:36:37
1
misho9000 :
Nice breakdown. There's also subsequent part that comes after the 14 positive and 4 negative possible actions are judged:
"The final scoring adjustments applied after Phoenix ML predictions and weighted score combination.
The *Author Diversity Scorer* attenuates scores for candidates from repeatedly appearing authors to ensure feed diversity.
The *OON Scorer*, applies calibration adjustments specific to out-of-network content. These scorers execute sequentially after the weighted scoring stage and before candidate selection."
2026-06-10 19:33:13
1
Kamikamihaaaa :
Hm I didn’t know that Twitter used to have PM’s boosting posts in the background.
2026-06-09 22:51:46
2
Montgolious III :
That's really interesting. The algorithm is remarkably crude. I always thought they'd be much smarter than that.
2026-06-11 08:12:14
0
Rasheed..._ :
summer is nice
2026-06-10 23:56:11
0
Manoj Chawla :
This is the comparison of fast food (you response/ taste/consumption) vs healthy food (is it good for you. Empty calories
2026-06-15 18:49:11
0
To see more videos from user @cirquedusunny, please go to the Tikwm
homepage.