😺 80% of jobs

PLUS: new text-to-video models

Welcome, humans.

Civilization-defining technologies: The steam engine, electricity and now, AI.

Here’s what’s going on in the world of AI today:

  • AI will be all over work 💼

  • Two new text-to-video models 📺

  • AI for doctors and games 🩺🎮

  • AI generates your travel itineraries ✈️

OpenAI: AI at Work Will Hit Nearly Everyone 👨‍💻

As chatter of AI automating away jobs begins to accelerate (see above), OpenAI made sure to let us know how many of us should be worried.

A new paper from OpenAI and the University of Pennsylvania claims 80% of the US workforce should prepare for at least some impact by AI models.

Key terms: for this study, "impact" means the human can do the same task in half the time by using AI. But it's still a human using the AI, not an AI completely taking over.

So who's safer from AI?

  • Jobs that rely heavily on science and critical thinking skills

  • Also: cooks, mechanics, dishwashers and athletes

Not as safe:

  • Jobs that rely heavily on writing and programming skills

  • Also: tax preparers, legal secretaries, journalists and writers 😬

The lesson: Most of us should get ready for more and more ChatGPTs and AI assistants in our lives. And remember to be nice in case they turn on you.


Come build an AI-generated feature film with us

If you loved Everything, Everywhere All At Once, pay attention:

AI is drastically lowering the barrier for film productions, so much that anyone with a camera, computer and OpenAI account can contribute to the next Hollywood movie. Now it’s your turn.

Say hello to White Mirror by CHAPTR, presented by The Culture DAO: they’re hosting a six-week challenge to create one of the world’s first AI feature films.

Their all-star team includes 10 top film creators and huge partnerships with OpenAI and Google Cloud to craft stories powered by synthetic media and generative AI.

And you could be their next team member…

If you join the White Mirror community, you can have a hand in shaping a feature film. Vote on ideas, give early feedback and discuss with creators every step of the way.

Become part of White Mirror on their Discord and help create the next blockbuster film.

Text-to-Video Heats Up 🎥

Generative AI is coming for every medium you can imagine. This week, it's text-to-video under the telescope.

Chinese tech giant Alibaba and AI video editing platform Runway both announced big text-to-video models (Alibaba, Runway) in the last few days.

These are the first text-to-video models that will be accessible to the public, contrasting Google's Imagen and Facebook's Make-A-Video, both still locked behind closed doors.

The quality doesn't come close to any regular video you see online. At best, they look like watercolors after an intensely vivid dream.

But they're coherent. And they show that creating a video is possible.

Next: Text-to-TikTok?

Around the Horn 🦄

  • "AI platform for film" software Filmustage raises $500K in funding.

  • Generative 3D builder Luma AI raises $20 million in their Series A.

  • Doctors rejoice. Microsoft is pairing its voice recognition AI (Nuance) and GPT-4 to generate clinical notes in seconds after a patient visit.

  • It's Game Developer Conference in San Francisco this week. Consensus: AI is going to massively disrupt video game development.

  • Google's ChatGPT competitor Bard is slipping beyond internal beta testing. Google Pixel "Superfans" are invited to try it out.

Treats To Try 🍿

  • HeadshotPro: Professional corporate headshots for $29/person using AI.

  • Double: Use AI to automate repetitive work on the Internet.

  • Tripnotes: AI-generated travel itineraries.

  • New model from Stable Diffusion automatically reimagines an image you upload.

DM us links on Twitter: @nonmayorpete & @noahedelman02.

Today's Top Pick 🥇

Disclaimer: we did not use AI to write today’s piece.

What'd you think of today's email?

Login or Subscribe to participate in polls.