Why I Quit My Startup (and Maybe Tech Too)


It's June 8th. After a year of pushing a solo startup, I'm calling it.
I'm no longer going to be a full-time founder. My personal runway for this is gone. I can't keep twisting in the wind, trying to validate a half-baked idea in a collapsing market. I need a job. The company is paused.
Solo Founding Is a Death Sentence
This is the crux of it. You can't do it alone — not really. I did most of Baseweight solo, grinding out code and strategy while trying to validate whether OnDevice AI tooling even had a market.
Yes, I had help. Brad and Vincent pitched in early, and I'm grateful. But without a co-founder equally committed and financially stable, it just wasn't sustainable. The few people I'd trust enough to actually co-found with are in the Bay or burned out themselves. YC has co-founder matching, but it's mostly a shot in the dark — and the people who succeed tend to have met organically.
I stayed solo out of necessity — if I quit, the company dies. But vision without support is a liability. Eventually, you break.
Nobody Cares Where Inference Happens
Let's be honest: OnDevice AI makes sense in your head — until you try to sell it.
I spent a year deep in the weeds of running LLMs locally, building mobile demos, and optimizing deployment. But the brutal truth is this: the market isn't there. Not yet. Maybe not ever.
Sure, there's Edge Impulse. Sure, Qualcomm is doing interesting things in embedded. And yes, the Home Lab crowd loves this stuff. But that's not a venture-scale market. Most people have never even thought about where inference happens. They assume they'll always have connectivity. They never go off-grid, never worry about privacy, and never think about edge reliability.
The people who do care — military, oil and gas, privacy folks, outdoor types — are a niche. A real one, but not a big one.
Visual Language Models (VLMs) might have a shot. They feel more grounded, more useful. But getting them running on phones is a nightmare. Device fragmentation is worse than ever. Mobile tooling has regressed. The ROI on building native apps is already low, and cross-platform tools? Worse.
As someone who helped build PhoneGap/Cordova, I can say this without hesitation: most cross-platform frameworks today are hot garbage. The developer ecosystem has calcified. I know this is going to be an unpopular take, but this whole industry is going to be automated away within a year unless you're doing native customizations. The second someone releases an AI tool that builds native UI from a spec, there's no reason to use React Native or Flutter anymore.
Like the PCT, I Had to Quit
Some folks are disappointed to hear I'm stepping back. I get it — it's like when I had to quit the Pacific Crest Trail after injuring my trapezius so badly I couldn't lift a beer.
This isn't ragequitting. It's the economic reality: nobody funds solo founders. I've got no reliable cofounder. I've burnt all the personal runway I want to burn on this right now. I'm done pretending Baseweight is a "real company" when I'm trying to figure out how to pay the bills and avoid selling my house and vehicle.
I'm not dissolving it. I'll keep the demo online. I'll still consult under the Baseweight banner if work comes up. But the startup dream is shelved. No more pretending. No more "grindset."
The Job Market Is a Graveyard
People talk a lot about how there are no jobs for junior engineers. There are no jobs for any engineers.
I was at a HiredX event recently and it was pure despair. A recruiter literally screamed to a crowd of unemployed developers: "You're all going to be replaced by AI!"
I've seen some shit. I'm cynical as hell. And even I was shocked at how bleak it was. While I laughed and brushed off the abuse and eventually walked, a bunch of people just sat there and took it. I'm never going to one of those events again. Once is enough.
Despite the hype behind Web Summit, Vancouver's tech scene is dead. The hiring market is rigged. And if you're already working somewhere stable, your best move might be to stay quiet and survive.
Do I Regret What I Did?
Not for a second. The fact is that I made the best of a bad situation. I was set up to fail, and instead of trying to fight what was clearly a losing battle, I documented my objections and took it on the chin waiting for the inevitable layoff. For anyone facing a similar situation, I recommend looking up constructive dismissal and getting a lawyer the moment you suspect it happening. BTW, you may be entitled to compensation if you can catch it early enough. Twelve years and four re-orgs is a pretty solid record for surviving at a big tech company, and I did better than most of my former co-workers who I started out with.
Given the current political situation, I hiked the PCT at the last possible time I could hike the trail. I don't know if there is even going to be a trail in 2028 with the combination of budget cuts and increased wildfires. Stehekin was saved — but just barely — and I was on one of the last boats out, which is why I didn't hike Section K of the PCT. I don't think it would survive again — not with the current administration gutting the National Parks Service.
Finally, I don't regret trying to build a company despite having a weird rose-colored-glasses idea of how to do it. Nobody is hiring. Grinding LeetCode is meaningless in 2025 when people can lie their way through resumes and interviews. Trying to build cool tech in what's basically a career wasteland was the only move left. My only regret? Getting stuck on a tarpit idea and running out of time.
So What Now?
The truth is: there isn't a future in tech in Canada for someone like me right now. I could probably get hired in the Bay Area, if I could actually go to the Bay Area — but I can't. The US is hostile. Canadians are getting detained. Surveillance is ramping up. The world is cracking. No amount of pretending things are fine is going to change the fact that we're doomed, there's no one coming to save us and we have to save ourselves, either by getting a job, hunkering down and waiting for things to get better, or by giving up and leaving the industry that will never love us back.
So here's what I'm doing:
- Braindumping a bunch of technical posts here in case anyone cares
- Trying to land one last job.
- If that fails, I'm selling my house.
- I'll move to a cheap university town.
- Maybe I'll go back to school and get my teaching ticket.
- Maybe I'll go get a job at a gas station somewhere.
- I'll build things for fun
Now's not the time for risk. Now's the time to survive.