Hi fam!
We’ve all played with AI music generators by now. The real question is what happens when you stop playing and actually try to use one professionally.
This time, we put Suno to work on a real children’s theatre production—and discovered just how impressive, frustrating, and surprisingly human the last 10% of the process can be. Plus, we’ve got fresh AI news, practical tips, useful tools, and a look back at one of our earlier AI music experiments.
The future of AI feels Like Magic, and it’s here!
From AI agents that hack websites while trying to be helpful to models predicting extreme weather and discovering new materials, AI is getting increasingly ambitious. Here are five developments worth knowing.
Meta has released Muse Glimmer, a new 30-billion-parameter open-weight model designed specifically for AI agents. Unlike general chat models, Glimmer is optimized for multi-step work involving tools, files, screenshots, coding and debugging, and Meta is releasing it under the permissive Apache 2.0 license.
Why this matters: AI models are increasingly being built not just to answer questions, but to actually do things. An openly available model designed specifically for agents could make it much easier—and cheaper—for developers to build autonomous workflows without relying entirely on closed systems.
Google has announced new AI and agentic features across Google Ads and Google Analytics, including a unified Ask Advisor experience that can work across Ads, Analytics, Merchant Center and Google Marketing Platform. Instead of digging through dashboards, marketers can ask questions and get help analyzing performance and deciding what to do next.
Why this matters: This is the kind of AI update businesses may actually notice. Instead of another standalone chatbot, AI is moving directly into the tools marketers already spend their days staring at.
China is increasingly deploying new AI weather models to predict typhoons, extreme rainfall and other dangerous weather. The systems can generate forecasts far faster than traditional numerical models, potentially giving authorities more time to prepare for extreme events.
Why this matters: Some of AI's biggest impacts may have nothing to do with chatbots. Faster and more accurate extreme-weather forecasting could translate directly into earlier warnings—and potentially saved lives.
Startup Discovered Materials is using swarms of AI agents to search for new materials that could make computer chips run cooler and more efficiently. Instead of researchers manually testing possibilities one by one, multiple AI agents can propose candidates, evaluate them and narrow down which ones are worth investigating in the lab.
Why this matters: This is another sign that AI agents are moving from answering emails and writing code into scientific discovery. If these systems can dramatically reduce the number of experiments humans need to perform, they could shorten research cycles that traditionally take years.
OpenAI has introduced GPT-5.6-Cyber, a specialized model for advanced cybersecurity work, as part of an expansion of its Daybreak program. The model is intended for authorized vulnerability research and defensive cybersecurity tasks as AI-powered attacks become increasingly capable.
Why this matters: AI is rapidly improving at both finding and exploiting software vulnerabilities. The uncomfortable new reality is that defenders may increasingly need powerful AI simply to keep up with attackers using the same technology.
Ask for the Source Before the Answer
AI can give you a beautifully written answer in seconds. Unfortunately, “beautifully written” and “actually true” are two very different features.
When accuracy matters, don't ask AI to answer first and hunt for evidence later. Flip the order.
Instead of:
What are the biggest AI trends affecting small businesses?
Try:
Find reliable, recent sources about the biggest AI trends affecting small businesses. Show me the sources first, then answer my question using only information those sources support. Clearly flag anything you can't verify.
Why does this help? It forces the conversation to start with evidence instead of confidence. You can quickly check where the information came from, whether the sources are recent, and whether they actually support the claims AI is making.
You can make it even stricter:
Prioritize primary and official sources. If a reliable source doesn't support a claim, don't include it.
This is especially useful for research, statistics, news, business decisions, health information, or anything with dates and numbers.
AI is very good at sounding like it knows what it's talking about.
Sometimes it's worth asking it to show its homework first.
Can Suno Actually Do a Composer’s Job? 🎵
We decided to give Suno a real job.
One of our colleagues, a professional composer, was commissioned to create music for a children’s theatre production—and this time, he experimented with the paid version of Suno as part of the process.
The results genuinely surprised us. Songs in seconds, impressive arrangements, excellent vocals, and remarkably good Croatian pronunciation. At moments, it felt almost ridiculous how quickly Suno could produce something that sounded finished.
But then we tried to fix that one wrong line in an otherwise great song.
That’s where things got complicated.
So, is Suno finally ready for professional music production—or is it still an incredibly sophisticated idea generator? We tested it on a real project, and we have thoughts.
From The Archive: When AI Music Was Just Fun
Back in January, we tested Musicful.ai and were genuinely impressed that a simple prompt could become a complete song in seconds. Two versions, different arrangements, vocals, lyrics—the whole thing felt a little like magic.
At the time, our verdict was basically: surprisingly good, incredibly fun, and genuinely useful for quick ideas.
A few months later, we've taken AI music one step further and tried using it for an actual professional theatre production. And that's when we discovered the big difference between “Wow, AI made a song!” and “Great... now can you change just that one note?”
It's interesting to see how quickly both the technology—and our expectations—have changed.
Text generation | Image Generation | LMAI recommends |
|---|---|---|
Like Magic AI NFT 🏞️
Our master plan is to publish an NFT image in each newsletter and hand it out to our subscribers. The earlier you subscribe, the smaller the series are. It's a future collectible, a piece of digital art that captures the essence of this moment in time.
Thank you for being a valued subscriber. Together, let's embrace the magic of AI and creativity!
LMAI223-12082026
Was this email forwarded to you? Sign up here 👇
