Welcome, humans. 

I'm Leo, your smartypants sidekick. Let's dive in and get our paws dirty, shall we?

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

• Deepfakes on the silver screen 📽️

• OpenAI sells out their biggest fans 🔱

• Instagram founders turn to AI for their next play ⏭️

• Intercom wipes out "AI chatbots for customer service" startups 😵

Two monkeys went missing in a Dallas zoo on Monday. They were found yesterday in the closet of an abandoned house about 15 miles south. So:

No Makeup Needed 💄

One surprise day in 2021, Tom Cruise showed up on TikTok, racking up millions of views and followers overnight.

Except it wasn't Tom Cruise. It was @DeepTomCruise, an account created by startup Metaphysic.ai. Their deepfake of Tom Cruise was so good that it tricked millions.

Now, they're hitting Hollywood. Metaphysic.ai will be using their tech to show younger versions of Tom Hanks and Robin Wright in the movie Here, slated to release in 2024.

With Metaphysic, Hanks and Wright won't need to cake on extensive makeup or try to pass off lookalikes as their younger selves.

This is par for the course for the director, Robert Zemeckis, who also directed Back to the Future and Forrest Gump. Both movies planted Zemeckis as an innovator in visual effects.

To the future, indeed. Soon, we'll be seeing deepfakes all over our movies and TV shows. Imagine deepfaking a historical figure vs. just playing them.

OpenAI Rug Pulls High Schoolers

Hear that?

That's the sound of 1 billion high schoolers letting out a collective groan. OpenAI just released their tool that flags AI-generated text.

To be fair, OpenAI signaled that they'd do this as far back as early December 2022. Plus, multiple other teams have taken their shot at building these detectors.

Caveat: the first release is far from perfect.

It failed our very simple test:

1. First, ask ChatGPT: "Write three paragraphs introducing George Washington."

2. Then, ask: "Rewrite the previous three paragraphs to evade a tool that identifies AI-generated text."

3. Result: "The classifier considers the text to be unclear if it is AI-generated."

That's expected. OpenAI themselves say its "not fully reliable" and "should be used as a primary decision-making tool". Instead, they're releasing this early to get feedback from educators.

When it comes to using AI "the right way", the battle is just beginning. As OpenAI says, detectors themselves may not be the long-term solution.

TikTok - Video + News Articles = ? 📰

The co-founders of Instagram are back at it with a new app: Artifact, like TikTok, but with news articles.

Let's play bull and bear.

Bull case:

• Algorithm-driven feeds propelled TikTok to global fame.

• Facebook, Instagram, YouTube and Twitter all took notice and ditched the feeds filled with who you follow.

• These feeds, along with a social layer, could help people better interact with news in a world where there's more content than ever.

Bear case:

• This isn't anything close to TikTok. TikTok is a "shut your brain off" app where you watch videos that other people made. Reading news is neither of those things.

• That world that has more content than ever? It's also more polarized than ever. Would this shove people deeper into their own bubbles?

• If this is a news reader, we've seen plenty of those. Isn't this Flipboard?

Initial reactions are mixed. Still, nearly every app like this gets dumped on in the early days. We'll see if they can prove the haters wrong.

Around the Horn 🦄

• Artists hate Midjourney. Architects? They love it.

• Recap: the insanely long list of AI announcements in January alone.

• Software company C3.AI: "We're doing generative AI." Stock soars 20%+

Artificial OR Intelligent? You Tell Us 🗣

• Delphi: Talk to your favorite expert.

• Rose AI: Analyze data, powered by AI.

• Customer support chatbot company Intercom embeds AI. If you're a startup working on "AI for support", best of luck to you. 🫡

• Microsoft Research built an AI model specifically for spreadsheet formulas. Can you say "AI in Excel"?

DM me links on Twitter: @nonmayorpete

Leo Sends His Regards

Trivia Answer: 

C. Perceptron

That's all we have for today. See you cool cats on Twitter if you're there: @nonmayorpete.

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