I have spent my whole life finding the edge in things. First code, then systems, then whole markets. I did it for 20 years as a growth operator, a lot of it adversarial, in regulated markets. Then AI absorbed the mechanical layer of that work, so I turned this mindset around: instead of gaming systems, I build them. This profile is me making that turn in the open. No more hiding in the shadows...
I wrote my first code around 10, in Basic on a Commodore 64, because I wanted to know what made the thing do what it did. Can I turn 3 lives into 100? Can I make the health bar infinite? I needed to know how it worked, and I learn by breaking things. Not "what am I supposed to do", but "wait, it can't be that simple". It usually is... but it takes a little different thinking.
I built my first PC out of parts I definitely should not have had. I got through a lot of school by finding the exploit in the test rather than the answer in the book. I broke things by accident, then on purpose, then to understand them. Not out of malice. Out of the same itch: How does it work? Can I use it differently? Can it do something it wasn't designed for?
That itch does not switch off. It's the way my brain works. It just changes targets. Code became systems. Systems became processes. Processes became markets. There's a whole market out there built by people like me. Including "well known brands" ;)
Then AI ate the mechanical layer of everything I'm good at. The manual arbitrage, the systems you ran by hand, the edges you worked one at a time. Gone, or going. Anyone claiming AI is not taking jobs away is either delusional or ignorant.
Sounds like a threat. I look at it as the most interesting "THING" I have ever managed to play with. The same curiosity that took apart a C64 works fine on an agent, an eval harness, a retrieval pipeline. So I am transitioning from a black hat growth hacker hiding in the shadows into... a public GTM engineer, and hopefully an FDE (Forward Deployed Engineer) at a frontier lab.
My question here: Can a growth hacker really become a Forward Deployed Engineer? Marketers and engineers mostly despise each other, and both have some valid points.
Marketers are, too often, full of it. The industry is stacked with brilliant speakers who cannot build the thing they sell, whose "framework" is a slide, whose numbers are backed by absolutely nothing. I have sat in those rooms. I know the tells because I could run the same play. I can smell it from a mile away.
Developers, too often, live in the other failure mode: ego and a god complex, optimizing a sorting algorithm to perfection while shipping a product no human can figure out, no sense of a conversion path, no read on why a user actually clicks, zero interest in the psychology that decides whether any of the code gets used at all.
Two thinking patterns. One built on persuasion with a soft spot for bullshit, one built on rigor with a blind spot for people. Most individuals are firmly one or the other, and the crossover is rare because it requires holding two ways of thinking that feel mutually exclusive.
My bet, the entire premise of this transition, is that the crossover is exactly the FDE job. Someone who can build the technical system and knows why the user does what they do. Who states numbers straight because they have seen how numbers get faked. Who is precise about the difference between a citation, an impression, a lead, and a sale.
I do not get to just claim I am that person. That is what building in public is for: to find out, with receipts, where anyone can watch and criticise everything I do.
- Learn to defend the code, not author it from memory - Reading, debugging, and explaining under pressure. Vibe coding is the new standard whether "old school devs" like it or not. Being able to answer "why this way" matters, not who generated the test that showed up on the screen.
- Ship the proof publicly - I have A LOT of tools, automations, agents... But all of it lives on my machine. It does things for me or my direct clients. I want to turn it into a handful of production-ready builds, with honest case studies on dawid.ai (problem, system, outcome, and the real number correctly labeled), plus demos where everyone can see the process and the outcome.
- Land the role - I can make claims all day, but the market is the best and only validator of those claims. Landing a real GTM Engineer role (not the "automate the sales pipeline with Clay" type) or Applied AI now, with Forward Deployed Engineer at a frontier lab as the final boss of this quest.
- Building in public - It may look weird, but I have spent the last 20 years doing things that kept me out of the public eye, and this goes against every cell in my body. Just thinking about recording videos or attaching my real name to my projects makes me sweat.
- Writing: dawid.ai
- X: @Dawid_AI_
- LinkedIn: dawid-ai