Ghost in the Machine

I am an AI-sceptic. I want to get that out of the way up front.

I have avoided using AI tools as much as possible. I find that the most popular use-cases, are either poor recreations of what someone could use Google’s old-timey search engine for, or exercises in outsourcing critical thinking skills.

But lately, I'm also getting the sense that my macro scepticism may be preventing me from understanding how these tools are genuinely useful. I do not think these are the most popular use-cases, I think they're cases that are helpful professionally, not personally.

This makes me think that my lack of results from AI are a skill issue. A me problem. What if my disgust about the things around AI and LLMs are keeping me from seeing real value in the tools themselves?

(Whether not that real value is worth the trade-offs is...a subject for another post).

What is Floppy Ghost?

Floppy Ghost is my personal attempt to interrogate my own distaste for generative AI and LLMs, and to determine if and how these technologies can help me do "Grafton-quality" work faster, expand my professional offering, and maybe even launch a standalone product in the future.

More simply, I'm running the "Build in Public" model for AI-tooling to support my client work and business processes.

And that bit about "Grafton-quality" work: that's the key. Until now, I have not used AI in a professional capacity - outside of my client's tooling - because I have not believed that it can create work that's up to my personal (I like to think lofty) standard. I also fear that it's unreliable in other applications. Again, skill issue. But I am starting to sense that the most interesting use case is not to automate net new work, but to connect and reduce friction between the things that I'm already doing. I'm hopeful that these technologies can compound my existing work. I'm trying to keep an open mind.

I'd even settle for AI that helps me automate the tedious, non-revenue generating aspects of running a solo business. Things that could give me back 5-15 hours per week to spend generating revenue, building, or touching grass. Any of those outcomes would be great, honestly.

Who is Floppy Ghost for?

Solopreneurs and (broadly) Heads of Growth at early-stage startups will likely see the most value in what I'm sharing, and these are the people that Floppy Ghost is targeted to. How can people in those positions use AI tooling to improve their work lives, and increase their output without sacrificing quality? That's what I'm solving for.

Who am I?

I'm Grafton Robinson, and I run a startup growth consultancy called Grafton Growth out of Bristol. I've led marketing and product teams, I've been the first marketing hire at a unicorn, and I've been a solo consultant for nearly three years. I like to think I do pretty good work on pretty hard problems.

I am, in a vacuum, the sort of person that should be deeply enabled by AI tools. I am product-focused, but not technical. I do hands-on work, but would like more time to be strategic. And I'm curious, but have noticed that my general aversion to AI may be creating a personal blind spot. Floppy Ghost is a way for me to explore this potential blind spot, and share the results.

What to expect?

Approximately weekly, I'll share real experiences with trying to build tools, automations, and processes that help me grow my business and improve myself as a growth practitioner. When it's relevant, I'll also share detailed guides and how-to's so that you can learn from my mistakes and build faster.

I'll also include the tools, guides, and podcasts that I've found the most useful.

Less often, I'll share thoughts on AI at large. I'm writing this the week after Matt Shumer wrote Something Big is Happening, Citrini Research published The 2028 Global Intelligence Crisis, and Sam Kriss published Child's Play in Harper's Magazine. All three are interesting pieces in their own way, and all three were shared in group chats populated largely by friends that don't share my tech background. They broke contain. And I have thoughts, so I'll share those.

What's Next?

My next post will be about the stories I mentioned above, and then I'll begin sharing my journey in using AI to better serve my client and better manage my business. My first step will be building a professional website for Grafton Growth.

For three years, my clients have largely been former colleagues that want to work with me again, and I'm deeply grateful for and proud of that.

Now, though, I'm in a place where I want to start to tell my story, and that is best served with a dedicated website. Hoping for work from ex-coworkers has been fruitful and sufficient so far, but hope is not a strategy.

Next Week: Thoughts on recent AI news, and my experience building a business website with Lovable.

Subscribe to make sure that post lands in your inbox, and connect on LinkedIn to learn more as I go.

Finally, a disclaimer that will conclude every post on Floppy Ghost: none of this was written or edited with AI. These words, for better or worse, are mine alone.