Hey Reader,

I’m writing this from my hotel room in Sydney after completing a big day of flying, debut webinarring, meetings, and then a mega board meeting to cap off my adventure.

Lots to cover this week, including:

• coaching versus AI coaching
• cloud AI versus private AI
• webinar recap - 3 things you didn’t know you can do with marketing and AI

So let’s make like a milkman from the olden days and deliver the goods!

-Marshy

Coaching versus AI coaching

Let’s get weird for a second.

One of the things that has tickled the back of my brain for a long time is the utility of my journalling.

I have blogged and braindumped since I was 17.

I have digital copies of things from that time, self-published a memoir at 29 (?!), have a steady and consistent “back catalogue” of braindumps from the last 10 years, and there’s not much time that goes by where I’m not clariying my thoughts repeatedly with writing.

I’ve always got value from this process.

It helps me think, it’s a powerful tool for gathering my thoughts, and there’s no question its assisted my mental health.

I was fascinated by the early note-taking tools like Evernote, impressed by the term “pensieve” in Harry Potter, and went deep on Tiago Forte’s Second Brain method for creating “systems” for capturing thought and helping me think better.

But I promised weird.

So then AI started coming and I thought to myself - there’s something in this.

One of the “early” movers in this space was Delphi - the idea being that there’s enough people out there with large online footprints (public or private) that want a way to interact with their thoughts (or share) them with others with what’s capable with AI.

I’d been obsessed with this idea - so to grab the founders attention I sent a big pledge through his Substack, got a call with him, and hashed out the idea:

​

They weren’t ready to absorb something like that at the time.

Dara was (quite correctly) targeting influencers.

They’re doing okay - and raised a Series A with Sequoia recently.

You could think of this as a form of “self-coaching”.

There’s also another product out with a similar style but focused on outcomes/outputs (and sounds a lot like coaching marketing - which is a terrible vortex to get stuck in I will add).

It’s called CoachVox.ai and here’s an influencer pimping it.

I feel like this kind of use of AI is more “personal trainer-esque” than the kind of coaching I respond to.

I was heartened by some of the comments too:

Maria the Master Coach chimed in with a great take:

Simon also articulated a heuristic that we need to be really aware of when engaging LLMs with personalised content:

It absolutely does tell you what you want to hear and is engineered to be appealing.

That’s not rational, dispassionate thinking (which is just as important for our fallible brains) and if we’re not critically thinking about this it’s a slippery slope.

For another way to go about these lines of thinking I suggest watching this video on the “master prompt” method that Tiago talks about with Hayden Miyamato.

This is clearly a growing area, and I have benefited immensely from working with my own human business coach.

The last video I wanted to share on this is a bit cringe I Replaced My $700/Hour Coach with NotebookLM – Here’s What Happened.

But I loved the take and distinction between human and AI coaches near the end:

Cloud AI versus private AI

The old war in nerd land was open-source software versus commercial software.

This war has waged since the dawn of computers, with the former embracing the indy hacker, release the software for the greater good ethos, and the latter typically represented as greedy corporations that is exploitative.

The lines appear to blur a bit when big corporations release open-source:

• Google releasing Android
• Meta releasing Llama

But in the game of high-stakes tech oligopolies - you can safely assume that these are more strategic positioning exercises than virtues.

In fact - Mark Zuckerberg has since back-pedalled on open-source amid safety concerns:

You can also very safely take interpret that as pure greed over what is being said.

But a new dynamic is starting to play out in the market.

People, organisations, institutions and governments are demanding more privacy.

The data held is too high-stakes to trust what big tech companies are telling us and now the market and capabilities are starting to catch up.

The emerging dynamic seems to be self-hosted AI versus cloud AI.

Here’s some examples:

• Locally in Australia, New Dialogue has started by promising private AI
• I’ve been partnering with another local company called Toothfairy AI that is built around open-source and giving control back to the organisation
• And OpenAI have released have released GPT-OSS, a private AI reliant on hardware instead of calls (Greg Eisenberg’s 1-sheeter is a good take on this)
• I also think some of n8n’s rapid rise is attributed to its ability to be self-hosted which draws big kudos from open-source advocates

​

Webinar recap - 3 things you didn’t know you can do with marketing and AI

I had a lot of fun with this.

It’s a different format and have been very eager to build out my own channel this way after doing a bazillion presentations with partners.

• Recording now up on YouTube​
• Slides here - and in the future they’ll be exclusively held in my new…
• ​Skool - I’m going to be dumping anything I think is useful in my community called GrowTechGood. I’ve been a beneficiary of this kind of approach and know some people love learning this way (and others drop off and that’s cool too) - I’m opening up as a freebie for early interested parties with the view to make it paid/closed down the road

I’m going to pump these out at pace - so feel free to sign-up at lukemarshall.net/webinar-wednesday

The next topic will likely be a solo deep dive on one of the topics today but am going to sleep on it.

–

It’s been a huge two weeks and looking forward to getting home for some downtime.

May you enjoy the rest of your week!

-Marshy