Here come the AI TikToks

There is a new social media app we need to have on our radar. OpenAI just launched the Sora app, powered by its new Sora 2 AI video generation model, and um…. wow.

How it works:

• The app creates/remixes astonishingly realistic AI-generated videos from a text prompt, with matching audio and sound effects.

• It centers around the notion of ‘cameos’ - a feature that let’s users record and use their likeness within the videos.

• Moreover, you can share your ‘cameo’ with friends or with everyone on the app to include/remix in their videos.

• You can view or remove any video that includes your likeness at any time, including drafts.

• Supposedly Open AI is “scaling up teams of human moderators to quickly review cases of bullying if they arise.” 🤔

• It’s invite-only right now, with the encouragement to “join with your friends.”

Here is a demo video by TikTok influencer iJustine so you can get a sense of what it’s like.

My initial take: it’s pretty mind-blowing… but also, it sure feels like it might unleash the floodgates around AI-generated social content…. and all that comes with it.

AI celebrities are officially here

If you haven’t heard of Xania Monet and Tilly Norwood you likely will soon. Each of them have either signed or are negotiating multi-million contracts with recording studios and Hollywood talent agencies. And they aren’t real people.

ChatGPT officially launches parental controls 

A highly necessary feature given the newly launched Sora app (which, lol, 99.9% of parents can’t get access to yet). But, at least it is a step.

To set up parental controls:

• Go to your Settings in ChatGPT and find ‘Parental Controls’

• From there, you’ll send an invite to your teen to connect accounts.

• Once your teen accepts, you can manage their settings from your own account.

• Your teen will also then automatically get content protections including reduced graphic content, viral challenges, sexual, romantic, or violent roleplay and extreme beauty ideals to help keep their experience more age-appropriate.

*Note that OpenAI has not yet launched the feature I described in a previous newsletter that would detect age based on chat behavior and turn on these protections automatically.

We are not as divided as we think 

I don’t know about you, but I’ve been really down lately about the state of our country and how angry and divided we all seem to be.

Well, I came across a body of research this week that gives me pause - and hope. The research studies the private opinions of Americans across a variety of topics (what we want for this country, what our views are on sensitive topics, how we view success, etc) as well as the disconnect between what we say publicly vs. what we believe privately. A few really interesting nuggets:

• There is a significant gap between what we say publicly vs what we believe privately: every single demographic group is misrepresenting their true opinions on multiple sensitive issues.

• We privately agree on most issues: for the majority of sensitive issues studied (43 of 66), ranging from abortion rights and school choice to legal immigration and voter ID requirements, 90% of demographic groups are privately on the same side of the issues.

• Moreover, Americans are profoundly wrong about how other Americans define success and the American Dream. We overwhelmingly believe success is about a meaningful life, not getting rich (we actually rank wealth toward the bottom). However, we believe that other people would rank being rich as the most important priority of all.

• Yet, right now, most of us are self-silencing. (63% of people have been self-silencing in the last year, with Gen Z, high income, and college grads self-silencing the most)

• Which becomes very problematic when data has shown ~10% of the people on social media (e.g. X) are generating 80% of the content and their views are more extreme / not representative of most of us. 1, 2

Key takeaway: being authentic and vulnerable and open with our friends and our communities has never been more important. And, it’s really good for us and our relationships. ❤️

Bring your child’s imagination to life

Speaking of mind-blowing AI video generation capabilities, Google’s Veo 3 model (available right in Gemini) hits the bar. You can simply type the video you want to see or upload a picture and it will bring it to life. The audio, sound-effects, and video quality are remarkable.

Something I’ve had a lot of fun with is uploading one of my boys drawings and asking Gemini to bring it to life. For example, Archer recently drew a character at school called “Avocadbro.”

• We went to gemini.google.com and selected “Create videos with Veo” in the tool selector at the bottom of the prompt window.

• We uploaded a photo of Archer’s drawing and asked it to “make this character come to life, dancing and singing“

And here’s the delightful (IMHO) output (turn up your volume for maximum enjoyment):

P.S. another very cool use case is to bring old photos of grandparents, etc. to life.

P.P.S. If you want to create longer films, you can do this in Google’s Flow tool

A Little Bit of Fun

Homeless invasion: Kids are texting their parents AI-generated images of a homeless stranger in their house, and parents are falling for it. 😂 😬