The Wisdom Wall

“And that's where we think automation has the highest leverage, not to replace people, but to really give the new workforce 10 times productivity, because we just don't have the people in America anymore to build these, you know, critical hardware for our customers as well as the DoD.”

“when you disconnect production from design, you get very bad at design.”

“The only way for Americans to be globally competitive with China is Literally just a labor equation, right? If we want to keep high wages in America, but compete globally, then we have to have high wage people that are using software and robotics to get a 10 X advantage. Otherwise you just end up in like a low labor…”

“large vertically integrated companies have two options. One is basically you do the whole thing and you win and you're like TSMC or like Tesla big, or you like die getting there. And there's no kind of halfway mark.”

“the way I would underwrite deep tech is basically like series B is the biggest risk point because you're starting to tip real money at the company. C and D is like what's already working. It's a fucking near monopoly. C to A is like, well, it's insane. And you're getting a lot of the company as an investor and, you…”

“Cause the other thing about deep tech is there is no, there is no iteration, honestly. Software companies, you can raise a couple million bucks, have an idea, test an idea, ship something new in four weeks. Whoops. Customer base sucks. Pivot, pivot, pivot, pivot, pivot. The amount of complexity in deep tech is just…”

“one of the reasons why the best companies have founded directly after recession has nothing to do with like pricing. It's just that you get people like Airbnb or Travis from Uber where they're just like, I have no other option. So my pain threshold is stupidly high because I have to do this with 200 grand. So the…”

“That's the reason why all ERPs are terrible is because you can't product manage your way to a good ERP. You kind of have to build it and use it yourself and then it finally becomes good.”

“The only way to produce that is that you run a manufacturing company with a vertically integrated software and robotics team so that you actually build the right product.”

“Many manufacturing companies have only picked like, you know, a low cost state. What ends up happening is No one wants to work there. So like looks good on an Excel spreadsheet, but, um, completely fails operationally.”

“the best engineers, frankly, just want two things. One is like three things. Work with other fantastic engineers. Work on really hard engineering problems that actually create business value and then have as few product managers as humanly possible.”

“The off switch can't be digital. And sometimes the best solution is just put an eight-bit drill through a hard drive.”

“I also think it's going to take a decade of sustained policy to go do it properly. So it has to survive multiple terms of the sustained tariff or subsidy policy to really reinvigorate it.”

“we're a capex-heavy business. So we like to spend debt instead of venture dollars on factory equipment.”

“many operationally heavy companies fail at this at scale because they end up having 2000 employees and 500 software engineers in their operations team and a bunch of product managers in the middle where You scope something out. It takes nine months to build it. Operations aren't really involved, so they don't adopt it.…”

“at some point, your success is the velocity of the org, not your own personal velocity, which is a transition that you have to make around 60 to 80 people.”