Ghost Hunting

9 Haunting Patterns That Could Define 2027

Author

Jim Louderback
January 06, 2025

This Week:  Terms of Service SNAFUs, Scammy Honey, 15 Icks that we’ll be talking about next year, and much more!

SCAMMY HONEY IS NOTHING NEW

Turns out Honey – the browser plug-in owned by PayPal and promoted by creators actually rips creators off by intercepting and stealing affiliate attributions  – and payments. The affiliate industry turned a blind eye to Honey’s blatant last click rip-off for years before last week’s “expose”.  Instead of banning Honey, affiliates tacitly endorsed the steal – because they made money along with Honey.  But their craven behavior is bad for everyone.  Attribution theft poisons performance data – and garbage sales data pushes Honey creators and publishers (wrongly) into the garbage.  Shame on the industry for ignoring this rip-off for so long.  As usually those who can least afford it – creators – suffer the most.  Want to fix this?  Retailers must block Honey from their affiliate networks and creators should focus on brand deals with traditional affiliate links to assure proper credit.

READ THE TOS

This company just got the right to hijack content from over 300,000 LinkedIn creators forever via an extremely harsh TOS.  I posted about it late on 12/31.  The CEO responded quickly, saying it was a last-minute stunt, and that they reused the TOS from their main product, and all that “perpetual sublicense” stuff wasn’t intended.  Intended or not, it’s now contractual fact.  Two lessons here: First, when it comes to letting a company scan your content, read the terms of service first.  Whatever the CEO says, they now have extremely favorable rights forever.  And if you’re doing a stunt for fun and a little publicity, don’t just reuse that TOS you slaved over for your main product.  Someone somewhere is going to read it.  And they might (like me), post about it too (the comments are fire).

SPONSOR:  Sixteenth Talent Misha Brown is a multi-hyphenate powerhouse: Creator, comedian, actor, musician, and host of a chart-topping Wondery podcast nominated for Best Comedy by The Podcast Academy. With over 4M followers and a USA Today nod as a “TikTok Comedian to Watch,” Misha’s sharp, heartfelt humor inspires worldwide. Catch him on a recent episode of the It’s No Fluke podcast by the Shorty Awards.

SIGNALS FOR 2025

I hope you enjoyed my last two 2025 trend-focused newsletters – the first on AI trends and the second on 25 Top Trends for 2025. 

In today’s part 3, we explore signals that COULD turn into trends in 2026 and beyond.  What’s a signal?  It’s a weird head scratcher, an uncomfortable story or a cringey quibi.   And if that irksome thingie keeps popping up then it’s a signal that something might be happening.  Often, it’s nothing.  But sometimes.  Sometimes.  

To that end, here are 9 signals I’m tracking, and six more on my radar.

  • Putting the Bot in Followbots: Meta just got roasted for dabbling in fake AI users, which made a lot of people quite uncomfortable. But I fully expect more AI constructs to be roaming through social platforms, following, liking, subscribing and viewing.  What does this mean?  It’s partly a continuation of the “Sub for Sub” or “Like for Like” from the early YouTube days.  And it’s nothing new. LinkedIn is already awash in commentbots, other platforms too.  But will an AI generated comment generate the same dopamine squirt as one from a human being?  I’m going to go out on a limb and say yes.  And if so, bring on the squirts.  Particularly if they’re nice.  Let’s face it, AI bots will probably deliver more inspiring, informational and funny comments than most of those I see today. 

  • Drinking With VTubers:  Creators who wrap themselves in 3D avatars are not only sliding into your YouTube and Twitch streams, they’re also infiltrating IRL.  In Japan you can walk into a bar, grab a drink, and cozy up to a VTuber.  They’re human-powered today, but soon they will be fully AI.  The potential trend here:  AI generated agents/avatars/vtubers will move beyond Instagram and Facebook and into real life.  I’m looking forward to adopting my favorite Pokemon as my familiar, and I expect Harry Potter fans will embrace snowy owls too.

  • Healthy AI Companions:  As AI spreads, we’ll see businesses built around the concept of “healthy AI companions”.  AI familiars will embody behavior and personality modification narratives that could treat psychological and neurological disorders.  Forget writing in your diary, as you’ll be chatting with your clinically trained persistent AI buddy all day instead.

  • My AI Gets Me: A friend of mine has a long LA commute.  But no podcasts, books on tape or sports radio for him.  Instead, he talks to ChatGPT. For hours.  Last time I saw him he insisted ChatGPT understands him better than his wife.  He was exaggerating.  I think.  However, AI will make the miles go by faster – and that could spell disaster for all the other things we do in the car.  Except sex, of course.  At least for now.

  • Mathematicians are Toast:  OpenAi’s new 03 model promises lots of change, but it might make that Math PhD less valuable than an English major. Why? Because math operates in a symbolic space, mostly divorced from the real world.  Thus it’s very susceptible to an LLM disruption.  Not tomorrow, but I wouldn’t suggest math as a viable career for the college graduating class of 2029.  As a Math Major myself, this one really hurts.  Will Bryk speculates about this and more here.

  • AI Every Minute: We’re starting to see a creepy rise of AI addicts – those who use AI services up to 100 times an hour.  VC Tomasz Tunguz is either on the forefront of a massive 2026 trend – or just another creepy VC.  I know Tomasz, so I think he’s on to something.  Interestingly Tunguz has become so reliant on his computer overlords that he’s already forgetting programming syntax that he used to master.  That’s the creepy part to me – when we start losing skills because AI is just so much better at it.  But creepy today = normal tomorrow.  

  • Driverless Taxis:  Normal in San Francisco, but visitors still get freaked out by cab sans cabbie.  It’s will take a while before headless taxis work everywhere, but creepy will slowly turn into commonplace.  And steering wheels will go the way of the buggy whip.  

  • Moore’s Law Was Child’s Play: Gordon Moore famously postulated that computer chips double in capacity every two years at a constant price.  But that law tapped out as the laws of physics took over.  But there’s a new law that replaces 2 with 1,000.  My old friend (and decidedly non-creepy VC) @Ashu Garg says that with the debut of Meta’s Llama 3.2, the cost of machine intelligence has dropped from $60 to 6 cents per million tokens over the past three years. That’s a drop of 1000x. Will it continue? Check back with me in December 2027.

  • Create-Predict-Win!  Prediction markets are a little creepy, the idea that if enough people bet on something it will happen – see Trump’s presidential victory in the US, for example.  Fans investing in their favorite creators still makes me a little uncomfortable.  But imagine them together in a creator-based mashup of predicting creator success while at the same time sharing in the success of those creators. While still gambling, oddly the two together seem less icky than each alone.  Keep an eye on this space.

More creepers on the horizon:

  • Creators move from houses and clubs to actual quasi-sovereign countries combining creators AND their fans.  

  • Nextgen scratch and sniff, as creators engage fans with more than sight, sound and motion.

  • The AI agent version of OnlyFans that must be coming soon (try combining this with the previous signal).

  • The end of mass media as AI creates ads, programming, and other content uniquely tailored for each person on the planet. 

  • The end of social as we all shift to private decentralized spaces and ignore the crowds.

  • AI moves from mainframe-style factories to micromodels for everyone – and an AI-enhanced resurrection of bittorrent (retorrent?) enables ripping, remixing, stealing and sharing of AI models globally. 

QUIBIS

YOUTUBE

  • AI Unleashed:  YouTube rolls out tools to let creators enable up to 18 specific AI platforms to train on their content.  The list includes A121 Labs, Adobe, Amazon, Anthropic, Apple, ByteDance, Cohere, IBM, Meta, Microsoft, Midjourney, Nvidia, OpenAI, Perplexity, Pika Labs, Runway, Stability AI and xAI.   Interesting to see IBM on the list.  I wonder what the OG Big Blue is up to?

  • Confessions of a Clickbaiter: It’s about time. I started using clickbait headlines and thumbnails back in 2008 to build up Diggnation and other Revision3 channels.  Finally, YouTube realizes there’s a problem!  Seriously, they’ve been fighting this problem for years – maybe AI will make it better, but don’t hold your breath.

  • BattleShop:  Big red partners with Shopee to dethrone TikTok in SE Asia.

META

TIKTOK

OTHER CREATOR ECONOMY

CREATOR TECH – AI, WEB3, VR, MORE

AI Disclosure: 100% written by me – no human or AI ghostwriters were involved (except for the cover art!).

Like this free newsletter?  Buy me a coffee and say thanks!  Or let’s do a meetup in your town.

I’ve built and sold multiple creator economy startups to top media companies – including Discovery and Paramount. Subscribe here on LinkedIn to get this newsletter every Monday.

Let me know what you think – email me at [email protected]. Thanks for reading and see you around the internet.