The end of Cruise isn't just the end of a particular robotaxi program; it's the end of a 10-year phase in the 75+ year quest for autonomous mechanized transport.

On December 10th, 2024, GM publicly announced that it would no longer fund the development of Cruise's robotaxi business, and instead fold its technical team into GM's driver assistance organizations.  While some saw this as a tragic mistake (Cruise’s founder and former CEO Kyle Vogt described GM leadership as "a bunch of dummies”), others took the opportunity to dunk on the program. Still, the overall sentiment was despair for the end of an era.

But whether you see the end of Cruise as a tragic mistake or a foregone conclusion, what's particularly interesting to me is how Cruise symbolized a period of great hope for autonomy, hope that barely existed before its rise and somehow exists even less today.

When Kyle Vogt first founded Cruise, VCs didn’t believe in autonomy. Kyle was not a one-time founder.  He also wasn't the sort of founder whose prior company was a rounding error success -- he co-founded what became Twitch, which was a VC home run when it was acquired for ~$700m in 2011.  Spurred by a lifelong desire to build self-driving cars, he founded Cruise in 2013.

But it wasn’t an obvious bet to investors.

Despite being one of the most successful founders of his era, Kyle went back to YC for Cruise.  Rather than being able to wink at a Tier 1 VC to raise $200m at idea-stage, he raised a $15m Series A from a fund with significantly less pedigree than you'd expect.

Because the big funds didn't believe in autonomy, unlike today, there wasn't an ecosystem of investors writing 9-figure checks for idea-stage companies that "could be big one day." And, frankly, when those funds reached out to robotics experts to diligence autonomy startups like Cruise, then pitching an easy attachment that would add Level 2 autonomy to certain models of Audis, those experts mostly said it would be decades before a robotaxi could hit the streets of a major city.

While the experts were saying that, the media was spinning a very different narrative.  Adam Jonas, an Analyst at Morgan Stanley, was talking about how robotaxis would be a $10 trillion market in 5-10 years. Travis Kalanick, still in vogue, was exclaiming how red-hot Uber could be disrupted if robotaxis exclusively went to a different rideshare platform.  Multiple people did the math that if people's cars only move 4% of the time, the car industry could potentially lose 96% of its sales.

And it isn't hard to see why that would sound both exciting and terrifying to General Motors (GM).  In 2007, sales of new vehicles decreased by 6% in a single year, which nearly led to the end of the storied company and needing a $11b bailout from the federal government.  It's not just that GM acutely had experience with near-failure; they also have been long in the tooth having lost 70% of their US market share since 1960.  Autonomy could either finish the job in wiping them out or be GM's path back to dominance.

That variance is a great incentive for business leaders to believe in unprecedented robot demos and optimistic claims to buy Cruise for $550m.  Six months later, Travis Kalanick announced a similar acquisition of 9-month-old OTTO for $750m, and the VC market freaked out.

‍

After OTTO's acquisition I've heard tell that every Tier I VC firm had an emergency Monday morning partner meeting.  Most of them had passed on Cruise, and none of them even got the chance to invest in self-funded OTTO.  When just one weird autonomy acquisition had happened, it looked like a fluke.  But two?  Two was a pattern that someone knew something they didn't.

Even if their robotics experts were saying that robotaxis were a long way off, somebody inside GM and Uber clearly thought that autonomy could be ready now. Whereas previously, VCs had rejected autonomy projects based on 1st Order Principles (hearing experts say it was harder than the founder said), these funds started leaning in based on 2nd Order Principles -- assuming that someone else knew something they didn't.

Over the next 5 years, VCs poured over $100b into on-road autonomous driving companies.  Sometimes, they found leading engineers in large programs and funded them without a deck; other times, they invested in sensor and infrastructure companies, hoping to sell pickaxes and shovels.  By 2019, Sand Hill Road had just about had its fill, but the frothy SPAC market of 2020 pushed many of those deals onto the PE and public markets.

At one point in 2021, there was over $150b of market cap for autonomous vehicles between stand-alone public companies and the attributed market cap for divisions within larger companies (such as Cruise's implied $30b valuation within GM).  This was even though, at the time, only Waymo was regularly driving on public roads without a person in the vehicle.  Somehow, the pre-product autonomy market was more valuable than the economy of Hawaii.

And then, it wasn't.

In 2022, Embark, after SPACing at a $5.6b valuation, failed when one of its key investors liquidated its position to make a capital call.  Similar fates befell TuSimple and Argo AI.  Cruise made an incredible amount of progress after the peak of 2021 interest -- deploying a fleet of hundreds of driver-out vehicles on the streets of San Francisco throughout 2022 and 2023.  But a poorly managed PR response to a highly visible accident and a home market that had long since lost its love for tech, led to it winding down driver-out operations in October of 2023.  And this month, the plug was pulled for good.

So, where does autonomy go from here?

The biggest heartbreak from the winddown of on-road autonomy the last 2 years is just how close we've gotten to a technology that truly has the potential to save lives and improve economies.  Multiple companies have driven millions of driver-out miles in major cities; multiple companies have driven driver-out trucks on public highways; we've achieved more in the last 10 years than the skeptics of 2012 thought we'd achieve in 25.

And it's not just robotaxis that have suffered.  Flying car companies, some of whom I thought, in 2018, would be the "next Cruise," have also flamed out.  Funding for new robotics companies has fallen through the floor, with the notable exception of humanoid companies whose claims so far outstrip conventional wisdom as to make Cruise's doubters seem optimistic.

The robotics industry is indisputably in the trough of disillusionment -- it is hard to imagine the market becoming more disillusioned without the failure of Waymo, Zoox, or Boston Dynamics to top it off.

But, if we're at the bottom, that means the market can only go up (however so slowly). Which means we must be entering the slope of enlightenment.  (Or, you know, Gartner is wrong).

As a phase in the hype cycle, the slope of enlightenment is a bummer.  It's almost always a slower buildup than the hype period, and it's typically characterized by markets starting to truly understand what technologies can and can't do, with early adopters seeing earlier benefits.  The robotics market seems to be seeing these early wins not on city streets but on off-highway use cases where the market demand has been so great that large traditional companies have been willing to try out hyper-limited autonomy -- systems that follow invisible rails between a set number of waypoints or drive a repeating pattern over a space.  In the next 2-4 years, we'll stop seeing robots as a "cool tech of the future" to fund, but closer to how companies think about buying new compressors or factory equipment: pieces of their technology mix that need to just work.

More than any company, Cruise inspired much of the development of robotics over the last 10 years.  While it is a shame that you won't be able to ride one around town, the autonomy wars they helped kick off have a lot of swords that, I hope, will start being used as plowshares.

‍