Welcome, readers. This week there is a surprising lack of YouTube updates. But there are updates for TikTok, Bluesky, Patreon, StreamElements, and lots of new AI features.
🔴 Join me live on Sunday to chat about this week’s updates
✅ What do you know about live streaming platforms? Take this week’s quiz.
👂Listen to the Creator Weekly podcast in YouTube Music.
Top news and updates this week
US TikTok had a bad first week, and some creators are trying alternatives like UpScrolled and Skylight.
Cheaper Google AI Plus subscriptions are fully global, giving access to some (but not all) Google’s AI-powered creative tools.
Experimental Project Genie can generate 3D worlds in real time.
YouTube is using Gemini in its recommendation system.
YouTube is making search safer for teens.
Patreon requires all users to move to subscription billing. They blame Apple.
The Edits video editing app has new features to copy fonts and other elements between projects and easier to access timeline settings.
StreamElements is raising money to “Keep it Free”.
Spotify now lets you direct Message groups with up to 10 people.
Google Search is making it easier to jump from AI Overviews to AI Mode.
Google Search is exploring new controls for website owners to opt out of generative AI features.
Meta announced their plans for 2026, including more AI in everything.
Bluesky shared their 2025 recap and plans for 2026, including post drafts, improved discovery and longer videos.
X maybe will add a warning on manipulated AI images. Or maybe not. There is just a vague Musk post.
There are new Gemini AI features in Google Photos and Chrome. Plus experiments that show AI-powered recommendations are not reproducible, and a new AI-labeling law in South Korea.
New Tips and Tutorials
If you are creating videos with Google Vids try their new animated caption options.
US TikTok Has A Difficult First Week
There have been many complaints about the new US TikTok, including bad suggestions in the new “for you” feed (trained on US data) and videos about recent protests in the US not getting views.
Some creators are getting notices of corrections for “inauthentic view counts”. Is that intentional? Or is it migration pains?
TikTok says a power outage at a data center “caused a cascading systems failure”, which could explain the issues. As of this writing (January 30th), the latest update was on January 27th, which says “made significant progress in recovering our U.S. infrastructure with our U.S. data center partner”. But there are still issues.
Many US creators are deleting the app and looking at alternative platforms.
UpScrolled claims “every voice gets equal power. No shadowbans. No algorithmic games. No pay-to-play favoritism.” It’s been described as a cross between TikTok and Instagram. In the past few weeks it has grown from 40,000 to over a million users (with 500,000 sign-ups over the weekend of January 23), as US users look for a TikTok alternative. The Australian company was founded by Issam Hijazi, a Palestinian-Jordanian-Australian technologist. It will be interesting to see if it grows beyond this.
TikTok alternative Skylight can also stream videos from Bluesky because of its AT Protocol integration.
And some people are using Loops.video which federates with platforms using the ActivityPub protocol, like Mastodon, PixelFed (created by the same developer) and Peer Tube.
And Substack had to chime in, with co-founder Hamish McKenzie declaring “The best insurance against censorship and cultural coercion is to build direct relationships.” Which is why the platform happily hosts literal Nazis.
Google Expands Cheaper AI Plus Subscriptions
Google AI Plus subscriptions are now available in 35 additional countries (including the US). The subscription includes 200GB of storage, plus expanded access to Google’s AI features (but less access than AI Pro and AI Ultra subscriptions). Existing Google One Premium 2TB subscribers in the countries where Google AI Plus is available will automatically get access to Google AI Plus as well. See the feature details. If you are using Google’s AI creative tools, it includes:
Flow: 200 monthly AI credits (Free accounts include 180 AI credits)
Gemini: “More” image generation and editing with Nano Banana Pro
Gemini: Veo 3.1 Fast video generation model access.
Google Vids: “Help me generate video”
Google Photos: “Limited” remix generations and photo-to-video generations.
Google Slides: Gemini features, including generating slides, custom images, infographics; remove background from an image; refine text.
NotebookLM: Higher limits for generating Audio and Video Overviews, Infographics and Slides. Increased notebook size with up to 100 sources per notebook (Free accounts can use 50 sources).
Create Your Own Worlds
Google AI Ultra subscribers in the U.S. can now access Project Genie, an experimental research prototype that lets users “create, explore and remix their own interactive worlds” using prompts from text or your own images. As you move around and interact with the 3D world, it generates the path in real time.
This is early days, but apparently the potential to “disrupt how video games have been made for over a decade” was enough to cause the stock prices of video game companies (including Unity Software, Roblox and Take-Two Interactive) to drop significantly.
How YouTube Uses Gemini for Recommendations
“Teaching Gemini to Speak on YouTube” is an interesting and not overly technical talk about how YouTube is using Gemini for its recommendation systems.
The system takes into account video title, description, transcript and info from sampled frames. Plus user demographics, and watch and engagement history.
The conclusion: "I think you're going to see recommendation and generative content start to come together in the future, where we're going to be recommending a personalized version of a piece of content, and in the future instead of recommending content we may even start creating it."
(This is not too different from Mark Zuckerberg’s vision of AI that he discussed in Meta’s recent earnings call. Read on for more details about that.)
Patreon Moving All Users to Subscription Billing
Patreon says that Apple has made changes that will force all creators to use subscription billing, rather than first-of-the-month or per-creation billing (which are used by 4% of users). Subscription billing charges patrons from the day they join, and then monthly on that day of the month.
US fans can still bypass Apple app store fees by allowing fans to join on the mobile web.
Patreon says they will try to make the transition easier with new features such as annual memberships, automatic repricing over time, and offering free trials and complimentary memberships. The deadline for change is November 1, 2026.
StreamElements Asks for Money
StreamElements is a platform that offers streamer tools like alerts, widgets and overlays. It also has a Sponsorship marketplace for Twitch streamers. One of their big selling points is that the tools are all “free forever”. But it sounds like that may not have been a sound business plan.
This week they set up a tip jar, which asks for donations to help them cover monthly costs (“Keep it Free”), with money going to server infrastructure, and other streaming tool costs.
As of this writing they have 47 supporters (out of more than 2 million users), who have donated $3,273, against their $300,000 monthly costs. This doesn’t seem like it’s going to really help.
Why now? They say, “With the current business climate and rising infrastructure expenses, we want to be more transparent about what it takes to power millions of streams while staying free and independent. This gives the community a way to chip in and help us stay independent.”
While they are committed to being “free forever”, if they fully run out of money they could end up being acquired by a company that isn’t so generous, or shutting down completely.
Video Creator and Live Streaming Updates
YouTube Shorts has a new AI-powered Remix feature that lets you add an object.
YouTube added new protections for teens that prevent search results from appearing for search queries linked to age-restricted or violative content. More information about teen safety.
There are new features for the Edits video editing app: copy and paste fonts and clips from one project to another (iOS), timeline settings that include toggle guides, frames and how edits are applied, apply edits made to your main track to other clips, and some previously launched features are now available for Android.
StreamYard added @ mentions to the in-studio chat. This shows a prominent notification to the mentioned participant.
Spotify launched Messages (private messaging for content sharing) in August 2025. This week they added group sharing with up to 10 people. The focus on in-app private messaging is definitely a trend, with recent updates on Instagram, Threads and TikTok, and YouTube experimenting with the option.
Web Publishers and Search
Google Search is updating AI Overviews. It uses Gemini 3 as the new default model. The big change is that you can “easily ask a follow-up question right from an AI Overview, and jump into a conversational back and forth with AI Mode.” Google says there are “prominent links to keep exploring”, although that’s not really obvious to me from the graphics Google shared. This is available globally on mobile.
In response to possible new requirements for Google Search by the UK Competition and Markets Authority (CMA), Google says they are “exploring updates to our controls to let sites specifically opt out of Search generative AI features.” They note, “Any new controls need to avoid breaking Search in a way that leads to a fragmented or confusing experience for people.” There is no indication when this could be available.
Meta announced their plans for 2026. Yes, it’s all about AI. Some highlights:
Nearly 10% of daily Reels views come from content made with the Edits editing app. It’s not just repurposed TikToks and YouTube clips.
AI dubbing of Reels will expand to more languages. It’s currently available in nine languages.
AI replacing creators? This seems to be a recurring dream for social platforms. During Meta’s latest earnings call, Mark Zuckerberg said: "Soon we'll see an explosion of new media formats that are more immersive and interactive and only possible because of advances in Al, our feeds will become more interactive overall. Today, our apps feel like algorithms that recommend content. Soon, you'll open our apps and you'll have an Al that understands you and also happens to be able to show you great content or even generate great personalized content for you." (Emphasis added)
Bluesky shared their predictions for 2026, future plans, and 2025 Transparency Report, noting “The best way to predict the future is to help build it.”
Bluesky’s users grew from 25 million to 41 million users, and have been working to scale their abuse systems. They promise appeals are reviewed by humans, even if the takedown was by an automated system.
People will post more during live sports and “you had to be there” events. They are making Feeds more central to the app, with new curation tools for live events.
People will click more links (“the bloodstream of the internet that carry us to all its marvelous places.”)
People will scroll less and rather use the platform more intentionally. They are working on improved “who to follow” suggestions and discovery, including topic tags.
People will think more before they post. New features they are working on include drafts, longer videos, more photos per post and easier-to-create threads.
There will be more integration with other Atmosphere-based apps.
X may be adding content warnings on manipulated images. There’s a vague post by Elon Musk, a reshare of an anonymous account mentioning the new feature. Musk’s post says “Edited visuals warning”, which is mighty vague. How will this work? There are no details.
More AI Updates and Tips
Google Chrome with Gemini 3 has a new “auto browse” agent that can browse the web for you, shop and even fill out forms and sign in to your accounts (using Google Password Manager). It will pause and get your confirmation for some tasks like making a purchase or posting on social media. This is rolling out in the US to AI Pro and Ultra subscribers. Other new features include an AI side panel, editing photos or creating infographics from web pages, and recommendations based on your connected Google apps and Personal Intelligence.
Google Photos updated its AI image-to-video to allow you to use your own custom prompts, rather than being limited to “Subtle Movement” and “I’m feeling lucky” options. And all photo-to-video creations may now include audio. Generated videos can be edited and “refined”. This is available to 18+ users. Learn more.
VSCO has added the ability to edit photos with a text prompt, including removing or adding objects, and adjusting levels for contrast and other settings. This requires a VSCO Pro subscription.
There’s an interesting study from Rand Fishkin of SparkToro that looked at reproducibility of AI recommendations. Real people entered prompts in Claude, Gemini and ChatGPT hundreds of times, ranging from recommendations for kitchen knives, cancer hospitals on the West Coast, science fiction novels, and Los Angeles Volvo dealerships. And the results were different almost every time. (Do read the article!) What this means for businesses and website owners is that there is no good way to see how you are ranking in the AI results. What it means for people trying to find information is that you may need to run your prompts multiple times to get a true sense of what top recommendations are, and even then it’s unclear what the recommendations are based on.
South Korea has a new law that forces companies to label AI-generated content. Clearly fake output, such as cartoons, can have an invisible watermark, but realistic images must have a visible label. AI systems for medical diagnosis, loan approvals and other “high-impact” uses must document how decisions are made. And “extremely powerful AI models” must publish safety reports. Depending on who you ask, it either goes too far, or not far enough.
Thanks for reading! 🌼

Social Media