China continues to clamp down on AI innovation.

If their tech stocks are any indication, it won’t end well.

🤖 Top Stories

🤯 Elon splurges for $250mm worth of GPUs

Elon dropped a cool quarter-billion on some flashy new GPUs.

The job? Generative AI for Twitter.

Ironic, given Musk’s call to stop AI development.

It’s not exactly clear what Twitter could use generative AI for. Summarizing threads? Fact-checking? Generating ads?

What’s clear is why they bought them versus renting. If this is a long-term investment that Twitter wants, it’ll be far cheaper to buy outright.

Especially as they recently hired two researchers from DeepMind to lead AI work.

They’ve got the pieces of a great AI product - team, data, and GPUs - but we won’t hold our breath for what they do with it.

😬 Universal demands people stop using AI to make music

AI-generated music is a real thorn in the side of the music industry.

How bad is it? Universal Music Group asked streaming services to stop artists from uploading it.

They’ve been sending takedown requests to streaming services left and right, according to a source. UMG wants Apple/Spotify to cut off access to music for any account that uploads AI-generate tunes.

The problem isn’t always in the end result of an AI song.

It’s what the AI was trained on. You can ask an AI music service to generate a song “in the style of” your favorite artist, and it’ll do that.

Using that artist’s music as training data.

UMG might have to go to the source - the apps used to train these models - to make any progress. A few websites that do this have already shut down due to legal challenges.

And it’s not like they can cut off all access to music - as we learned with Napster.

Not a great place to be.

We expect streaming services to develop their own solutions for detecting AI music.

👀 China, of course, has rules for AI

While the US considers AI regulations, China pushes full steam ahead.

The Cyberspace Administration of China released a draft law outlining rules for developing AI.

Spoiler: it’s totalitarian.

AI-generated content must “reflect core socialist values, and must not contain content on subversion of state power”.

It must also not contain “other content that may disrupt economic and social order.”

Every new AI product has to go through the review, if it’s signed into law. They’re calling this a "security assessment."

We’d liken it more to a content filter. Instead of security from attacks from bad actors, it’s protection from the “wrong” information by users.

It’s China, so this will zoom through the government.

Let’s hope the US doesn’t take any notes.

😎 AI tools, tips, links

1. ChatGPT runs your Python scripts. Learn more

2. Let AI do your jobs on the web for you. Learn more

3. Virtual photo studio for any two people. Any. Learn more

4. ChatGPT for Google Sheets. Learn more

5. Robots sorting trash in offices. Learn more

GPT-6 when?

From Reddit

🤖 Got feedback for us?

Head over to our Twitter, @RobotRockDaily, and let us know what you’d like to see more or less of.