Good morning.

Some Monday motivation for you. On a Saturday.

Now for some news.

🤖 Top Stories

👀 Meta heats up the AI wars

You’re probably thinking “Ok, we’ve got big AI announcements from Amazon, Microsoft, and Google.”

Why hasn’t our favorite artificial intelligence, Mark Zuckerburg, said anything?

Well, wait no longer.

Meta finally jumped into the AI ring.

Meet LLaMA, a new Large Language Model for researchers. From Meta.

Zuckerburg has higher ambitions than only generating text.

From his post on Facebook:

I know what you’re thinking.

Hold your Llam-… er, horses.

Meta isn’t adding ChatGPT to Facebook.

Yet.

This is for AI researchers. Part of the appeal is that this model has several “tiers’ of parameters to choose from. They range from 7 million to 65 million. Generally, more is better, but more expensive to operate. ChatGPT has 175 million, for example.

Wherever this takes Meta, it will 100% work out better than their Metaverse mess.

😬 Microsoft teaches ChatGPT to…teach robots

We’ve moved past humans teaching robots.

We’re not on to humans teaching robots teaching robots.

Microsoft released a report detailing how they used ChatGPT to control a drone.

Spoiler: It worked really well.

And in case you’re wondering, yeah. That scared them, too.

So much so that they cautioned AGAINST doing this without careful analysis.

Microsoft used ChatGPT to command a drone to perform simple tasks in a house setting. Like locate an object or take a photo.

ChatGPT’s ability to transform natural language into instructions made it a breeze.

Microsoft imagines a day when you can talk to a machine in plain English with instructions. Something like "heat up my lunch.” That machine will understand and execute.

The sky is the limit with this. What’s to stop a country from letting a chatbot power its drones?

Or, more importantly, what’s to stop my oven from perfectly cooking my frozen pizza?

😭 Uh oh: AI art is NOT protected from copyright

My dreams of being an artist are over.

The U.S. copyright office just ruled in a case for an AI-generated comic book.

It ruled that AI-generated imagery is NOT protected under U.S. copyright law.

So…that’s a huge blow.

For what it’s worth, the court did recognize that effort might have gone in to making it.

Just, well, not quite enough.

Not only is this tough for people trying to make a living in AI-generated imagery.

It’s also big for companies trying to make their moats in the same space.

One thing’s for sure: I wouldn’t stake my company on the copyright ownership of AI-generated ANYTHING.

This also isn’t the last we’ve heard of this issue.

The courts will have to read up on Midjourney, and QUICK.

😎 Cool links 

1. AI health coach.

2. Fashion shots.

3. Quick AI workflows.

4. ChatGPT for product design.

5. List of generative AI startups.