Hello, geeks!
Good news! Meta has found a way to offer better user support.
Bad news! It’s with an AI bot, which probably won’t be able to get you your wrongly suspended account back.
Oh, and Zuck is planning to switch from human moderators to AI, too. Lord help us! 🤖
That’s all in the same week that Meta has cut back even further on the metaverse that gave it its name.
Details about all of that coming up. Plus the all the latest news, virals, Oscar memes, and new platform features spotted in the wild this week.
But first, don’t miss these 👇
2 cool things…
🔥 You don’t need 10 tools to run 1 social campaign. I found this one that does it all. [ad]
-💬 This viral "LinkedIn Speak" translator turns any text into cringe business thought-leadership BS. 💀
__________
P.S… Geekout’s exclusive interview with Instagram Edits’ product boss Matt Simari is being recorded next week.
I will be sending you a special edition of Geekout featuring the interview in full…
…Including a sneak peek at a NEW FEATURE coming soon to Edits!
SPONSORED ▶︎ TOP MARKETING TOOL
You shouldn’t need 10 tools to run 1 social campaign 😭
Your content's performing. But your tech tools stack is a mess. 😬
You're posting consistently. Engagement looks good.
But your leads? They're slipping through the cracks.
Follow-ups are manual.
Or worse… forgotten entirely.
And everything lives across five different tools that don’t talk to each other.
Sound familiar?
That’s the real problem.
HighLevel fixes it 👇
One place for your funnels, CRM, automations, email, and client delivery.
Everything actually connected.
No more duct-taping six apps together.
No more “why didn’t that fire?” moments.
Here’s what changes:
📥 Capture leads straight from your content
🔁 Automate follow-ups before they go cold
📊 Track every deal (no spreadsheet chaos)
🧰 Manage clients without juggling logins
💵 Create predictable, repeatable revenue
It’s the system you thought you had… until now.
Most businesses wait months before fixing this.
👉 Try it for FREE for 30-day — before more leads slip through 🚀

Whistleblowers allege Meta and TikTok weakened safety controls during the race to build more engaging algorithms [RACING]
Instagram will discontinue end-to-end encryption for DMs [DITCHING]
Meta reportedly plans to lay off 20% of its workforce as AI development costs grow [DOWNSIZING]
Meta has been caught failing to stop illegal financial ads in the UK [SLACKING]

The metaverse dream really is dead
We’ve been here before - saying RIP to Mark Zuckerberg’s metaverse dream. But each time, it seems more convincing.
This week Meta announced plans to shut down Horizon Worlds, its 3D social world, on VR headsets this June. The platform would still be available on mobile, but the VR version - the clearest articulation of the ‘metaverse’ that the company has - would be gone.
Meta swiftly rowed back a little after user outcry. The VR version will still be live but not receive anything but essential updates. Still, to all intents and purposes, Meta will no longer be focused on a ‘metaverse’ in any meaningful way.
The news follows recent scaling back and refocusing of Meta’s Reality Labs division, which spearheaded the metaverse’s development.
The news spurred blunt headlines this week like: RIP Metaverse, an $80 Billion Dumpster Fire Nobody Wanted.
A more charitable view would be that some Silicon Valley bets work out and some don’t, and at least Meta has plenty of in-house AI tech to fall back on. This week’s Geekout has lots of examples of Meta rolling out more and more AI - it’s undoubtedly the company’s future.
But Mark Zuckerberg’s metaverse vision stood out as something his counterparts at other big tech companies weren’t doing. Even Apple’s Vision Pro (similarly sidelined of late, it seems) is focused on utility rather than virtual worlds.
Zuck would probably say there’s still time for the metaverse to make a comeback when the time is right. And VR is still important for Meta. The company said this week that Quest VR hit record usage in 2025 as 100 apps passed $1 million in revenue.
But for the foreseeable future, the failure of the metaverse dream is a reminder that just because big tech companies tell us something is going to be big doesn’t mean it really will be.
It just means they need something to tell the stock market that quarter and they’re willing to spend a LOT of money on it, even if it ends up going nowhere.

Meta
/ Facebook
Meta swiftly reversed plans to shut down Horizon Worlds on VR headsets after users and creators revolted. Instead it’s going into long-term ‘maintenance mode’ so Meta can focus on the platform’s mobile version
Meta is replacing Sponsored labels with smaller Ad tags on Instagram posts, and testing the same on Facebook
Meta is updating its $800 Ray-Ban Display HUD glasses with widgets, a Calendar app, and live captions for phone calls
Meta says Quest VR hit record usage in 2025 as 100 apps passed $1 million in revenue
Meta announced its NYC store will become a permanent flagship
Meta’s Horizon Store is adding better app discovery, deals sections, and monetisation tools
Meta tightened up Moltbook’s user rules after acquiring the social platform for AI bots
Facebook has a new monetisation programme to attract popular creators from TikTok and YouTube
Facebook has a new option to create a shorter version of a Reel that links to the original longer version.
Facebook added tools help creators quickly detect and report impersonators stealing their content
Facebook has been spotted more widely rolling out the ‘Collage’ Story type
__________
Instagram
Instagram is finally rolling out tap-to-pause on Reels (what took them so long?!)
Instagram is rolling out voice effects for DMs to more users
Instagram added an in-app shortcut to its TV app
Instagram Edits has added new features including five new cinematic effects, camera upgrades, and more
Instagram is widely rolling out the ability to trim a Reel before sharing it as a story
Instagram has been spotted offering a new transparent navigation bar on Android
__________
Threads
/
WhatsApp
Threads now lets you share your posts to Facebook Stories
WhatsApp is working on a feature to automatically translate messages across 21 different languages
WhatsApp is testing guest chats, letting users message people without a WhatsApp account
WhatsApp is rolling out a new profile tab to iOS users
WhatsApp is testing a Status redesign modelled on Instagram’s Stories interface
WhatsApp has added the ability to access Ad Preferences directly from the Status tab
WhatsApp is testing the ability to link a companion device and access WhatsApp using your Facebook account
WhatsApp is working on built‑in noise cancellation for voice and video calling
WhatsApp is working on the ability to edit polls
WhatsApp is testing a new preview for shared Threads links
WhatsApp is testing a new way to manage media on desktop
WhatsApp is coming to select Garmin smartwatches
WhatsApp’s beta app has a new icon
__________
X / Twitter
X teased a possible new dislike button on replies. Users can flag replies as spam, misleading, “AI-generated,” and more
X on the web is limiting profile previews, encouraging you to view them in the app
xAI is being sued by US teens who allege Grok made sexual images of them as children
Closing arguments took place in the class-action trial over Elon Musk’s $44B Twitter acquisition tweets
….and Elon Musk and SEC are in talks to settle a separate lawsuit over the 2022 Twitter deal
X is complying with the EU as it appeals its €120 million fine over its verification badge and transparency
X might have to make Grok less ‘spicy’, as Musk’s tactic of blaming users for AI sex images could be foiled by EU law
__________
TikTok
/
YouTube
TikTok and Tubi are partnering to help TikTok stars create original long-form shows for the streaming platform
TikTok launched an interactive BTS ARIRANG experience for the band’s comeback
A US senator is looking for answers to questions about the reported $10 billion payment the Trump administration could get for the TikTok US deal
TikTok is officially rolling out #BookTok bestseller charts across the UK and Europe
TikTok launched new ‘Prime Time’ and ‘Logo Takeover’ ad formats
TikTok is testing the ability to share a comment as a text post
TikTok now displays stories posted by each artist in the audio pages
TikTok is encouraging users to use the TikTok Studio app, offering multiple benefits as incentives
TikTok introduced an option to link posts to the TikTok Studio app
TikTok has added a new advertising objective in TikTok Promote, ‘More likes & comments’
TikTok is rolling out a new animation for the templates search section that suggests trending templates
TikTok appears to be planning to add a new, private Watch Later folder in the Collections tab
TikTok parent ByteDance has suspended the global rollout of its new AI video generator Seedance 2.0 amid legal concerns from Hollywood
…while US senators have demanded ByteDance “immediately shut down” Seedance 2.0
__________
YouTube
YouTube has become FIFA’s ‘Preferred Platform’ for the 2026 World Cup across all devices. Broadcast partners can stream the first 10 mins of every match and select full games, on YouTube
YouTube introduced a pop-up asking users to rate videos as “AI slop” or not
YouTube is tackling misleading thumbnails with instant video previews
YouTube’s head of music has pledged to protect artists’ rights, likeness, and creative control, as AI reshapes the music industry
YouTube increased the thumbnail upload limit 25×, from 2MB to 50MB
YouTube has been criticised by a child safety group Fairplay for investing $1M in AI kids media company Animaj
__________
LinkedIn
/
Bluesky
/
Telegram

LinkedIn’s new game: Patches
LinkedIn expanded its games lineup with a new puzzle game, ‘Patches’
LinkedIn now lets advertisers use data from the platform to target professionals watching TV
LinkedIn expanded creator ads and video tools for marketers
The rebooted Digg’s open beta has shut down after just two months, blaming AI bot spam. The platform is being rebuilt from scratch
Bluesky now lets creators of (good) bot accounts label their accounts as such
Bluesky has revealed that it raised $100 million in new in investment last April
Telegram now supports text formatting, like Quote, Spoiler, Bold, Italic, Monospace, Link, and more
Google expanded its Reddit integration in “What People Are Saying” results
Linktree launched a YouTube playlist carousel that auto-updates with new uploads

Meta’s Manus has launched a desktop app called ‘My Computer’, to turn your Mac into an AI agent.
A bit like OpenClaw, it enables agents to manage files, run commands, and automate workflows.
Play safely, folks!
__________
Also spotted this week…
Meta AI will now include more real-time global news, entertainment, and lifestyle content in responses thanks to partnerships with News Corp, Le Figaro and more
Facebook and Instagram now have a 24/7 AI-powered support assistant
…but it probably can’t get you your account back (booooo!)
Meta is working with Signal creator Moxie Marlinspike to add an encryption layer to Meta AI chatbot interactions
Meta officially rolled out new Manus AI integrations for marketers and creators on Ads manager, WhatsApp, and Instagram
Meta is now promoting Manus AI as an “AI work partner” for Facebook Pages
X is testing ‘AI content detected’ alerts before users repost content.
X added a new “Summarize” feature for articles published on the platform. You can also get Grok to read the article aloud
X moved the Grok model selector in the chat message box for a cleaner UI
X upgraded Grok’s AI video generator, letting users create clips from multiple images
YouTube has launched an AI-powered ‘Reimagine’ option for Shorts
TikTok is testing AI-powered ‘cover titles’
Snapchat’s Easy Lens app now shows which prompts were used to create a lens and allows you to remix it to create a new one.
Adobe’s AI image generator can now be trained on your own art
BuzzFeed launched its own AI slop apps as it struggles to survive

Meta’s New Mexico trial entered its fifth week as explosive child safety claims mounted.
The state alleges systemic failures, citing abuse backlogs and weak enforcement systems…

What happens when the ‘king of memes’ is challenged to make a new meme live on TV?
Another iconic meme, obviously!
Conan O’Brien’s opening monologue at the Oscars on Sunday included a nod to Leonardo DiCaprio’s status as ‘king of memes’.
But could DiCaprio create a new meme on the fly? Of course he could! He rose to Conan’s challenge and a new classic was born.
__________
__________
__________
__________
Also:
21 brutally funny reactions to Meta’s metaverse failure

While Leo was busy making legit meme magic at the Oscars, viewers at home were busy faking photos of the night for engagement.
As Deadline highlighted, moments like this did GREAT on Instagram and X, even though they never actually happened.
That feeling you’re feeling? Reality slipping through our fingers 😬
Fake:
Fake:
__________
More viral news this week:
A flood of AI-generated war videos and images about Iran has spread widely across social media
Celebrities are jumping on the viral “What were you like in the ’90s?” trend on TikTok and Instagram

“Be careful who you trust for product recommendations: social media influencers are spreading brand-related misinformation in a much more toxic way than ordinary users, research has found.”
__________
More brand news and views:
Apple’s ‘brainrot’ strategy for marketing the MacBook Neo on TikTok is working
Advertisers aren’t happy about picking up Meta’s European tax tab
Brands are shifting toward smaller creators who drive higher engagement and conversions
Advertisers are shifting to conservative creators over progressives under Trump
Movie studios now release “trailers for trailers” to grab TikTok-era attention
Lay’s launched a celebrity-packed WhatsApp group chat for World Cup fans
Experts at SXSW 2026 say no new Alix Earle-style stars will emerge in 2026

"This is a memification and a gamification of war… It’s an appalling way to represent conflict."
__________
Also quoted this week:
“This is deeply disturbing. Under our laws, online platforms have a duty to keep people safe online. We're already tightening the rules on AI chatbots and I'm ready to take further action if necessary.”
- UK Prime Minister Keir Starmer responds to claims Meta and TikTok let harmful content rise after evidence that outrage drove engagement
__________
“The AI companies are claiming fair use, but this argument is bogus… It’s bogus because while they claim it’s fair to use the work of creators as training data, they do multimillion-dollar deals with rights holders and publishers like Disney and Condé Nast and Vox and Warner Music.
“If it’s legal to just use it, why pay?” Why pay them and not creators — not the millions of illustrators and musicians and writers — whose work has been consumed by these models to build hundreds of billions of dollars of value for these companies?”
- Patreon CEO Jack Conte calls out AI companies’ hypocrisy

Remember all the hype around Clubhouse?
For a hot minute in 2020, group audio in front of an audience was the next big thing…
Perhaps because we were all stuck indoors for weeks at a time and were desperate for social contact.
This podcast from The Verge explores the rapid rise and just as swift fall of a brief social hit.

Want to get into podcasting but feel overwhelmed?
Rebel Audio has launched an AI-powered all-in-one platform to simplify podcast creation
It handles recording, editing, publishing, and monetisation without switching tools. AI features include voice cloning, translation, transcripts, and content generation tools.
Read more about it at TechCrunch.

Yikes. This is a tad embarrassing for Meta…
A Meta AI agent exposed sensitive company and user data to unauthorised employees for nearly two hours.
Meta classified the security incident as ‘Sev 1’, the second-highest level of severity inside the company.
Meta says no data was misused… but the incident highlights the growing risks of autonomous AI systems as CEOs race to be the most AI-savvy.
__________
More WTF this week:
Meta Ray-Bans have a 'fart candle' prank problem
Royals and celebrities have been warned that lip-reading videos are exposing their private conversations online
Activists have plastered provocative posters across London accusing TikTok and Meta of fuelling a youth mental health crisis
The UK has banned an AI image editing app’s ad suggesting it could remove woman’s clothing
Fans are demanding an end to influencers hosting major red carpet events… because they’re not very good at interviews

Tumblr swiftly reversed a major change to the way it handled reblogs after users HATED it.
The new approach worked a bit like Twitter/X. Users could like, reply, and reblog individual posts in chains.
The update fragmented the iconic collapsed thread format into separate posts, killing what made Tumblr unique.
The company believes it still needs to revamp reblogs, but it’s going back to the drawing board.

“Google has pulled its AI feature that summarised crowdsourced health advice, stepping back from risky, non-expert guidance.”
Who could POSSIBLY have predicted that wouldn’t work out well?
__________
Also heading to the social media graveyard:
Snapchat is killing its standalone Bitmoji app on 15 April. Bitmoji avatars will remain available within Snapchat and partner integrations


Airbnb is looking for a TikTok Social Manager [London, UK]
__________
More social media jobs:
Remote needs a Director of Social [Remote, worldwide]
Circle is hiring a Head of Media [Remote, worldwide]
Shakespeare’s Globe is recruiting a Social Media Officer [Hybrid, London, UK]

Instagram and TikTok linked to worse mental health…
That’s according to the 2026 World Happiness Report published earlier this week.
The study shows WhatsApp, Facebook linked to higher life satisfaction due to social connection focus.
However, the report also found 1-hour daily use of social apps boosts wellbeing, more than 0 hours or excessive use.
But researchers warned average usage of social apps currently sits at near 2.5 hours per day…
… Or 13.5 hours if you’re a social media manager 😆

You probably have platform images sizes lodged in your head. But are they still correct?
Buffer has you covered with this in-depth 2026 image size guide for Instagram:
__________
More new pro-tips:
The best time to post on X in 2026 — based on data from 8.7 million posts
Meta, Snap, and YouTube insiders reveal common mistakes new creators make in 2026
The 11 best social media analytics tools for creators and marketers
How AI agents decide which brands to recommend: Trust is the new ranking factor
Pinterest shared tips on how to utilise its AI-powered ad options

$10 BILLION
More stats:
BOUGHT IT: Reddit offers a high 82% return on ad spend when Amazon purchases are accounted for, according to a new survey

Creator pay still skews heavily towards the biggest creators in 2026.
But brands are shifting toward smaller creators who drive higher engagement and conversions.
The segment is growing 10× faster than macro creators.
But earnings are uneven: the top 10% take 62% of all creator payments.
And nearly half of all US creators make under $10K a year.
More charts:
HERE COMES THE FEAR AGAIN: Creators fear algorithm changes more than burnout in 2026

__________
Also spotted this week…
The UK Government is exploring mandatory labels for AI-generated content to fight deepfakes. [Reuters $$$]
Australia’s eSafety regulator says child abuse material is “particularly systemic” on X. [The Guardian]
Adobe will pay a $150 million settlement over its really annoying subscription cancellation policies. [9to5 Mac]
China is stepping up scrutiny of Meta’s $2B Manus deal. [New York Times $$$]
Discord has open-sourced Osprey, its real-time rules engine used to detect harmful activity. [Discord blog]
Some creators don’t see immediate value in Instagram’s controversial ‘Shop the Look’ AI test. [DigiDay $$$]
Meta has shared how Facebook built ‘Friend Bubbles’ to power social discovery on Reels. [Engineering at Meta]
Meta will replace Nielsen DMA data with ComScore Markets for auto ad targeting. [Social Media Today]
Advertisers will soon be able to buy Reddit ads through Pacvue. [Adweek]
YouTube is expanding its live-stream “Jewels” gifting system to South Korea in March 2026. [Social Media Today]
Meta has signed a 5-year AI infrastructure deal worth up to $27B with Dutch cloud firm Nebius. [CNBC]
SerpApi asked a court to toss the Reddit scraping lawsuit it faces. [Search Engine Land]

_________
Also worth a read:
Long-form creators eye taking over TVs – and chasing bigger brand budgets
How Reddit CEO Steve Huffman got the upper hand with AI
The danger behind Meta’s decision to kill end-to-end encrypted Instagram DMs
How the Trump White House tries to sell war and death as a game
Why social search visibility is the next evolution of discoverability

I love reading about cringey messages from ‘influencers’ demanding special treatment from businesses…
Or flexing like billionaires with faked lavish lifestyles, when all they really have in the bank is thousands of duped followers.
So enjoy this list… if you can remove your fingers from over your eyes! 🫣
Look out for my tweet featured at #22 in the list… I exposed a travel influencer faking clouds (seriously!) in her Instagram-perfect holiday pics.

And there we go… you’re fully caught up and completely debriefed on everything new in social this week.
I’m heading off to finish final edits on a consulting report I completing for [REDACTED COMPANY NAME] all about AI products for marketers.
The deadline is today, but I think I’ll need another 2 days to wrap it up 😭 Wish me luck!

Gif by saito90 on Giphy
Goodbye, geeks!

P.S. This newsletter is edited by Martin SFP Bryant
































