TOGETHER WITH POPHATCH

Hello again, human brain, and welcome back to your daily munch of AI news.

Here’s what’s on the menu today:

• Sora 2 is here. But you’re not invited 🤳

  OpenAI launched its video model, Sora 2.

• AI just replaced your band 🎸

  Suno Studio is here. Like GarageBand, but AI music.

  New here? Subscribe! 😎

Peep today's Spot the AI at the bottom. 👇

MAIN COURSE

OpenAI says “go TikTok yourself” 🤳

Yesterday, OpenAI launched Sora 2. It’s an AI video generator with synced audio, a social app, and more.

What's new?

Sora 2 generates better AI videos with improved physics, synced audio, and can create 10-second clips. But the real change is the social app. Think TikTok meets AI video generation.

For more info on the first Sora, click here.

How do I get it?

You can't just sign up. OpenAI is rolling out invites gradually. Early users get friend passes, then ChatGPT Pro subscribers, then Plus and Team users. An iOS app launches tomorrow in the U.S. and Canada.

What can you do in the app?

Create videos from text prompts, “cameo” yourself or friends into videos (with consent), remix videos like TikTok duets, follow creators, and scroll a feed of AI-generated content.

Can I put myself in videos?

Yupp. The Cameos feature lets you capture your likeness and voice to insert into AI videos. It's opt-in only, and public figures can't be generated unless they upload their own cameo.

What about deepfakes?

OpenAI claims every video carries watermarks and metadata to prove its AI. They've also added parental controls for teen accounts. Will this really stop misuse? Probably not.

Who's Sora 2 competing with?

Google's Veo 3, which currently makes better AI videos. OpenAI is betting that the social features will make Sora 2 stand out from other video generators like Runway.

Beware. AI video is about to flood your feed.

SPONSORED BY POPHATCH

Your AI Cofounder 🤑

Someone else is building your idea right now. While you’re figuring out GTM strategy, they’re already three steps ahead. While you sleep, they’re spotting market opportunities you’re missing.

PopHatch is your AI cofounder that never sleeps and is literally obsessed with your startup’s success. Built specifically for people creating AI SaaS, PopHatch handles everything you don’t want to (or don’t know how to) do: market intelligence, competitive analysis, validation, GTM strategy, scaling decisions, and ongoing business operations.

This isn’t another AI tool. It’s your actual business partner. PopHatch is where you run your company. While you focus on building your AI SaaS, your AI cofounder is working 24/7 on making sure it succeeds in the market.

The result? You build faster, launch smarter, and scale without the guesswork. Your AI cofounder spots opportunities before your competitors do, handles the business strategy you hate dealing with, and never stops working toward your success.

Perfect for: Anyone building AI SaaS who wants a cofounder as obsessed with winning as they are, without giving up equity or dealing with human drama.

The urgency is real. Every day you’re building without business intelligence is a day your competitors are getting ahead.

SIDE SALAD

AI just replaced your band 🎸

Suno launched a new platform that lets you create/edit music with AI. Think GarageBand, but you can generate every instrument with a text prompt.

What is it?

A generative audio workstation where you can allegedly build a complete track without music knowledge. Hum a melody, type a prompt, and the AI generates it.

More on Suno, here.

How does it work?

Like a normal music editing platform, but you don't need pre-recorded audio or skills. You can generate drums, swap out vocals, adjust pitch and tempo, or tell it to “make the synth sound dreamy.” Then edit everything on a multitrack timeline.

What's the catch?

It costs $30/month for Premier subscribers—three times the price of Suno Pro. Also, it's desktop-only for now.

Is it any good?

I made a couple of songs and found them perfect on paper, but hollow. The music didn't feel alive until I started heavily editing. It’s useful for brainstorming. If you don’t have audio production skills, you’re not making any hits with this.

Is this legal?

Eh. Suno is facing lawsuits from major music labels for allegedly training its AI on copyrighted music from YouTube and other sources without permission.

Will this flood Spotify with AI slop?

Probably. Spotify already gets 100,000 uploads per day. If everyone can make music this easily, discoverability could collapse even further.

YOUR DAILY MUNCH

Think Piece 🧠

Should engineers feel threatened by AI? CNC programmers, data scientists, and technical writers face the highest AI exposure, while traditional engineers are safer.

Startup News 💰

Anthropic launched Claude Sonnet 4.5. It claims it’s the best coding model in the world.

Sakana AI unveiled ShinkaEvolve. It improves computer programs through evolution-style mutations, finding solutions way faster than traditional methods.

Research 👨‍🔬 

SciReasoner — an AI model trained to understand and reason across different scientific fields by translating between natural language and scientific data formats.

Seadream 4.0 — ByteDance's AI image generator that can create, edit, and combine images from text prompts—generating high-res images (up to 4K) in just 1.4 seconds.

MEME FOR DESSERT

SPOT THE AI

3 of these are real squirrels. 1 is AI. 🐿️

Which one is AI-generated? 👇

Ideas? Comments? Complaints?

Respond to this email or hit me up on 𝕏.

Until next time 🤖😋🧠