Post

Conversation

David Watson 🥑
Post your reply

As an aside, this chart is how every startup should be judged (well, cashflow vs. operating profit, but that's pedantic). Not valuation, not who got in at what price and out at what price, but how much more cash did you produce than what you consumed, and over what time period.
I had claude code me a FoodDeliveryApp and it hallucinated all restaurants, menu items, and drivers, but if i launch it then im going to be a BILLIONAIRE according to ex paramedic psychology major in LA and all the wall st retards who listened to him
You don’t get it. People want completely bespoke front ends that change every time they use them. Only AI allows that kind of mayhem.
This is another reason why you can't just "vibe code" an app like Uber. The infrastructure behind it and customer acquisition took a decade before profitability.
Uber helped create the driver network and those drivers drive for many platforms. AI agents can totally take advantage of that to increase driver pay and lower customer price.
Just in time to be replaced by robo taxis. Alternatively, if Tesla pivots to selling/leasing them to Uber, could cause profitability to go vertical.
Utilizing and coordinating humans at a reasonable cost is the hard part.
"i just can't play with bulls anymore"
Quote
Nate Lorenzen
@anatelorenzen
Software investors right now.
The media could not be played.
A marketplace is the most difficult business in the world. Change my mind. Either you’re lucky or you just have to burn a lot of cash to attract people to it
Note that the tiny profits they now have is due to raping the fuck out of their drivers, taking >50% of revenues in many cases. Absolutely terrible rent seeking behavior akin to the worst company town abuses.
So I agree… but the promise of an AI version of it is that you can lure drivers with promise of higher pay. The new company doesn’t need the same infrastructure and can run with lower margins The impact of AI will be cost structures and margins Uber can do it themselves though
How much of this was building both sides of the marketplace that agentic can now more easily replicate ? The chart starts in 2014. Today compute is faster and cheaper, supply and demand behaviors are trained and instrumented to engage with each other digitally. Not
I feel like the self driving car threat is the real one. Depends how good waymo is compared to others to see if commoditized
Well they fought all the hard regulatory fights. Plenty of space for competitors to come in with better pricing models for drivers.
They will see two years of profitability. Also there was no hard part. They were chosen by the powers that be and given a blank check to form a monopoly. See also: Amazon
There's islands in the med with their own uber style apps that are functional. The interface and technology doesn't seem to be the hard bit
Building the frontend is not the hard part of scaling a marketplace Anyone who thinks it is hasn't tried to solve the cold start problem Honestly the code is the easy part
The problem will be in a few years when there are three or four major AV providers. Then an AI agent on your phone can do some damage to Uber.