171–180 of 274 posts
The point being TiKTok does a ton of labelling of videos that allows the algorithm to “see” the content - or at least qualifies lots of elements of each video. And then optimises for joy and happiness responses from users.
Other useful observations in the podcast
Now ask the same question for YouTube shorts ;) Hah. Google is pushing shorts hard these days. Honestly, they need to give an option not to show shorts. In most cases, shorts are either from TikTok from the same content creator or re-uploaded by some random person. Even if I liked topic X on shorts, it never suggests similar shorts. It is hard to believe that Google cannot crack the correct algo to show interesting r…
Did you know that Youtube has a game streaming service like Twitch? Neither did I. They've hidden it so fucking well that even though they pay (tens of) millions for exclusive contracts, people can't find it. But shorts they keep pushing on every UI.
Although org wise it's under YouTube Gaming livestreams can be about anything. The option to go live is shown when you click the upload button. Live streams show up in your video feed and the top of your subscriptions. In explore and on the sidebar of every page is a link to the live metachannel. On game metachannels there is a tab to see people streaming the game. Livestreams show up in search and you can filter for only livestreams.
I think it's pretty hard to avoid since streams are promoted more compared to videos. You must just not watch creators who stream on YouTube or are interested in topics that lend themselves to being streamed.
A new platform is not built on a single day. Even creators found a new platform and decided the first thing they do should be re-posting their old works.
Earlier quoted context omitted.
The problem with asking "actual users" in this way, is that it fundamentally ties the user's pre-existing conception of these apps to their responses. If a user had been served the same content on Reels that they were on TikTok, it's difficult to see them having the same reaction.
An actual scientific experiment would have like recently fallen away Amish teens compare these apps having never used them.
the UI itself can be tested separately
Earlier quoted context omitted.
In the android App? Where? Could you please share.
When I am on the webpage on my desktop there is the x button that says something to the effect of "don't show shorts". I click that and the strip of shorts are removed. For the app shorts is a different tab on the bottom of the screen. It goes Home, Shorts Subscription, Library. Shorts are never mixed him to my regular feed.
You can click not interested on them all, it takes some time to get through them all, but after a week or two they are back and it's rimce and report.
Yes I do have a shorts tab at bottom as well. I never click it. It's the shorts in my feed that are a pain in the butt.
Now ask the same question for YouTube shorts ;) Hah. Google is pushing shorts hard these days. Honestly, they need to give an option not to show shorts. In most cases, shorts are either from TikTok from the same content creator or re-uploaded by some random person. Even if I liked topic X on shorts, it never suggests similar shorts. It is hard to believe that Google cannot crack the correct algo to show interesting r…
YouTube is plagued by unnecessarily padded content because I believe creators were rewarded depending on how much time they spent watching their videos (and hence number of ads that can be shown). Vast majority of YouTube videos I come across can be reduced by 90% in length without any loss in information you are seeking. YouTube needs to simply change incentives for the content creators and TikTok will have a massiv…
Earlier quoted context omitted.
> 1. The Reels algorithm is simplistic as fuck. I was scrolling through Instagram recently and a three reel “ad” came up. I accidentally tapped on the center one and it turned out to be a pimple popping video. Now every reel I see is that. No exceptions. I am sure if I actually watched more Reels I would get better content but that’s work. OMG this... a million times this :/. YouTube has similar issues--where you wat…
I’ve always wondered why they are so bad at learning such an obvious lesson from their supposed biggest threat. TikTok seems to always mix it up; even if you do actually want to watch pimple popping 24/7, they seem to throw in some totally unrelated thing once in a while. Turns out people easily find things they didn’t even know they would enjoy this way, and when they have many interests, their feeds are way more en…
I've tried to sign up for instagram, but it always said "email is already used", including for emails that I had just created. Maybe they're losing users because nobody can sign up? Or maybe the web sign up doesn't work, and you're only allowed to sign up with a facebook account (now they have your data) or the mobile app (they also have your data). If so that's the kind of abject dishonesty that might also lose you…
I like that they asked actual users, and provided actual examples; it's something I've always felt off about Instagram's Reels vs TikTok. However the takeaway is that your TikTok feed is curated by hand by artisan robots vs Instagram is curated by lizards who only care about data. Both companies are completely data driven with some amazing engineers; but why is the Reels product boring, while TikTok is engaging has t…
As a millennial I can tell you exactly why TikTok is better: 1. The Reels algorithm is simplistic as fuck. I was scrolling through Instagram recently and a three reel “ad” came up. I accidentally tapped on the center one and it turned out to be a pimple popping video. Now every reel I see is that. No exceptions. I am sure if I actually watched more Reels I would get better content but that’s work. 2. Reels are often…
Yes, you in fact get better recommendation as you use it more.
Reading HN, sometimes it feels like I must be the only one in the world enjoying good recommendations from both YouTube and Instagram. I do have many thousands of watches and likes on YouTube and Instagram, but it didn't feel like work, because I accumulated them over years of use.
If you thought Cambridge Analytica was bad, wait 'til you find out what a China-based company can do with detailed preference and arousal data from young people.