​

Hi readers, and welcome to another edition of AI versus Marshy - the newsletter that helps you keep up with developments in artificial intelligence without feeling left behind.

This week we look at:

• So what are Apple doing with AI?
• A cool niche - turning popular Tweets into videos
• Learning with tailored emotional feedback

Before we get into it - here’s Mr Fluffles, a character I drew last year:

And here’s Mr Fluffles by AI if he was writing this newsletter:

Alright - enough goofing around, there’s LOTS to cover!

So where is Apple in all of this AI business?

Via 9 to 5 Mac.

It’s pretty safe to say Apple, Amazon, and Google all shat the bed on voice recognition.

Whether it’s Siri, Alexa, or OK G - none of them work half as well as OpenAI’s Whisper can.

It’s fair to say Apple know this.

They also probably don’t agree with Ross -

Siri being improved to current levels of technology is a very quick changing of the game.

There’s a lot of hype about wearable AI (and I’ve covered a good chunk of it) - but do you know what’s easier than convincing people they need a new piece of hardware?

Adding more utility to the hardware they’re already addicted to.

This will likely move Apple from laggard status to serious competitor quickly.

I’d love to say “you heard it here first” but improving Siri is a no-brainer.

Turning popular Tweets into videos

Heard through Starter Story.

​Nathan Latka wrote a book called “How to Be a Capitalist Without Any Capital” and I think about this excerpt a lot:

I think this is the reason I love Repurpose Pie - there are a LOT of thought leaders and content creators churning out content and threads on X/Twitter.

The founder noticed that some of the Tweets that popped, could easily be repurposed into a video format with AI to be re-used either back on Twitter or elsewhere.

So she found some viral Tweets and started building her MVP.

The business now runs with a healthy email list, referrals, and content creators signing up for the recurring service.

​

You can learn more about what they do in this walkthrough video.

Now that’s a good pickaxe ⛏️

Emotionally-informed education.

Via Tom Osman.

What gets me really jazzed about no-code tools is the way they click and enable new use cases with technology that’s available today.

AI takes this same principle but blows it up to mind-blowing levels.

Tom (ex-Makerpad) shares an example:

Take learning something via LLM chat agent.

You get feedback, you ask clarifying questions, and get an answer about a concept you’re happy with.

Now what if it was also tracking your response with visual recognition that links expressions to your emotional response?

It could detect how well you’re absorbing the education and whether it’s resonating. It could then feed that information to the chat agent/teacher in a repeating loop.

Getting better and better at teaching you based on your emotional feedback…

🤯

The flex of this is quite sinister - what if such technology was used within a social media tool like TikTok, to keep you more and more engaged…

😬

The next decade is going to be murky at best.

–

That’s it for today!

I’m enjoying the projects that are going and am keen to keep building up the big fast value call approach - I’m very determined to up my coaching this year and will keep doing these for free until it’s no longer sustainable.

Take care and stay curious - we’ve got this!

-Marshy