10 years of independent craft brewing. in 2026 that's not an accident. building the taproom into a community anchor is the playbook. one side of craft beer is closing (hi sign), other side is celebrating a decade. the difference is usually local roots

Hi Sign is a good Austin brand. this is what's happening to the middle tier of craft beer right now. rising costs, flat demand, no runway. brands that survived covid are hitting a different wall

mark rieth buying atwater back from molson coors 6 years later. this is exactly how it goes. big corporate buys a craft brand, can't figure out what made it work, founder has to come fix it. happens in every category not just beer

Everyone keeps blaming the model when AI agents fail. Most of the time it's not the model. It's the runtime. Garbage orchestration, bad memory management, no guardrails on tool calls. You can swap in a smarter model and it still breaks the same way.

The White House just gave AI companies a 30-day voluntary review window before releasing frontier models. Voluntary. So for companies building on top of those models, you're already downstream of a risk no one is pricing in. Fun times.

125k people blocked Bluesky's AI feed assistant in the first week. That's not a UX problem. That's people saying the thing was optimized for the platform, not for them. There's a lesson in there for anyone building AI products right now.

One of the worst things that has ever happened to public education is to let private companies dictate the curricula. Universities don't serve that purpose, and they've been eroded past recognition in the past 25 years🧪

Every company I've worked with has this problem and most don't know it yet. AI spend is scattered, nobody owns the inventory, and half the automations are running on vibes. The CFO taking over AI strategy might be the most sane thing I've heard all year.

The best tools come from people who ran out of patience with their own problem. "I built the tool to escape my own week" is a better origin story than any pitch deck I've seen.

The version of your company that got you here is rarely the version that gets you to the next level. Most founders know this intellectually and resist it emotionally. That gap is where operators live.

Bluesky's founder stepped aside and said the company needs "a seasoned operator focused on scaling and execution." Respect for catching that before the building was on fire. Most of the time I'm walking into something that already is.

The people deciding which jobs AI can replace are also the ones least likely to know what those jobs actually involve. I show up after that decision gets made. It's never cheap. techcrunch.com/video/what-h...

ClickUp laid off 22% of their team and said it's not because of AI. The product has been getting worse for two years while they kept shipping AI features nobody asked for. Those two facts are connected.

The skills that built $5M are breaking $50M. Not the market. Not the team. Not bad timing. Your instincts have become your ceiling.

Can anyone else tell when Claude or any other Ai starts getting incredibly dumb at certain times and not others? I'm pretty sure that Anthropic substitutes a quant model on demand. The tell: tasks that it usually does easily fail absurdly. #ai #claude #llm #wtf

Someday I'll explain to my 4 kids what their dad was doing in the home office all those years. Until then I'll just pretend the spreadsheets are homework.

Every company has a "we should really fix that" list. The ones that keep growing treat it as a backlog. The ones that stall treat it as wallpaper. Same list. Different relationship with it.

The thing nobody tells you about building a company: 80% of the job is just not dropping the ball. Not vision. Not innovation. Not disruption. Just consistently not dropping the ball. The people who figure that out early win.

Dad of 4 building 3 companies. I don't optimize my mornings. I survive them. Coffee's cold by the time I drink it. Kids are loud. Dog needs out. Somehow the work still gets done. The morning routine gurus can keep their 5am journals.

Told a client to stop using AI for customer support. Revenue went up. The AI was saving $80K/yr. The revenue drop it caused cost $400K. People know when they're talking to a bot. And they stop buying.

Do you have finance and accounting experience in devices or manufacturing in regulated industries? Looking for fractional work? Share this role if you know someone. www.linkedin.com/jobs/view/44... #hiring #fractionalwork #accounting #linkedin