The Social Media AI-pocalypse
it's a tough time to be an influencer
If you’ve been keeping up with current events in the world of AI video generation, you know that we’ve had a dramatic turn of events over the last three weeks as ‘Sora level’ video is almost here. China’s Kling can almost match it, but only if you have a Chinese phone number. Luma-labs Dream Machine comes close and seems to have fixed their ten hour wait times. Pixverse.ai is almost there, and certainly the best value, because it just works consistently. And Leonardo.ai is pretty good at converting your generated images to video clips, but has no direct text to video prompting.
However, RunwayML broke the sound barrier this week when it released demos from it’s Gen3 model which are truly ‘Sora level’ and (rumored) to be available today. While I have bashed on Runway in the past for disappointing results, they were clearly playing the long game and now have a product I am willing to pay for. If this rate of improvement keeps up, we’ll have a huge range of Sora competitors by the end of the summer. While it certainly isn’t won’t be good enough for feature films, I can think of a few applications with an an immediate use case. Television commercials and social media content.
Your typical 30 second ad, while occasionally brilliant, is often formulaic. Great writing sets them apart, but the visuals need to be familiar for easy recognition. Between Midjourney, RunwayML (when I get it) and an upscaler like Topaz Labs ($300), I think I could make a professional grade TV commercial in the very near future.
The biggest impact will be felt in social media, as aspiring influencers can now create perfect life content with little or no barrier to entry. The results are still a little janky around the edges, but pretty soon you’ll be able to craft scenes of the perfect mimosa brunch, the perfect Hawaiian getaway, and the perfect outfit. These cases involve using the external tools I described, but the technology is rapidly being developed for in app use. Snapchat will have an AI assistant that lets you dress your avatar in AI generated outfits by the end of the year.
TikTok just launched Symphony, a new suite of creative solutions powered by generative AI. This new suite is designed to make content creation easier than ever for marketers and creators alike, simplifying the creative process while scaling it to new dimensions.
TikTok Symphony is designed to elevate your content creation journey every step of the way. Blending human imagination with AI-powered efficiency, TikTok Symphony enables businesses of all sizes, creators, and agencies to level the playing field, boost productivity, and uncover valuable insights. From scripting to video production to asset optimization, Symphony transforms the complex into the simple—if you can imagine it, you can create it.
And this, from Shelly Palmer’s newsletter
Former Snap engineer Bobby Murphy has launched a new social network called Butterflies, which integrates AI-driven entities that can engage in conversations, provide companionship, and assist with various tasks. The platform's AI companions are designed to learn from user interactions to continuously improve their responses and adapt to individual preferences, creating a personalized and dynamic social experience that evolves with the user.
Here’s a link to a longer article about this product
Why it matters:
We’re about to be flooded with AI generated video content on our social feeds. Some people will label it appropriately, but many will not. For the untrained eye, it will be hard to distinguish what is real and what’s fake. If you’re a content creator, the value of your work, and your potential income, are about to become drastically watered down. Why pay an influencer when you can generate an avatar to pitch your product? We’re currently on the cusp of AI generated, animated talking heads selling us things.
Thomas Smith has a great essay and a series of examples that I drew from.
And, for your amusement: Original IG post here
Quickies
Yes, I’ve been hating on Adobe lately, but we’re all still using the software and we still need to generate pdfs. Did you know that you can now put AI generated images in your pdf? Check it out
UX inspiration of the week. Here’s a speculative project that tries to imagine the next phase of user interface design, where one master AI-powered interface could organize everything we do on our device. Check it out
Tengr.ai is yet another free-ish Midjourney competitor with a browser based interface. Easy to use and get started with. Read about it. | Try it
Midjourney now has a feature called ‘style reference’ or —sref for short. Once you’ve generated and saved a style, you can share it with others. Christie C. is an excellent Midjourney author and she’s graciously shared ten of these styles. So if you want to up your Midjourney game, check it out
And one last tip to start your deep fake career. I haven’t tried it yet, but Hedra looks like the best solution for lip synch animation so far. Read about it
And then, moments before I was about to hit ‘publish’, a bunch of new Kling videos were released. Check it out
I hope you’re having a great summer and learning lots of exciting new things. Remember, if you notice that your friends are jealous of how ‘in the know’ you are, you can always tell them to subscribe. Send me a message, ask me a question, keep in touch.