Remember when going public was supposed to be the finish line? For Rivian, the 2021 IPO feels like a lifetime ago. The stock launched at $78, hit $179 in a matter of days, and everyone thought this was the next Tesla killer. Fast forward to today and shares are hovering around $17. That’s an 80%-plus wipeout. Ouch.
Yet something curious is happening. Even with deliveries sliding and the entire EV market catching a cold, Rivian stock is actually up more than 30% this year. Why? Because investors are no longer betting on how many trucks Rivian can build in Normal, Illinois. They’re betting on something much bigger: whether Rivian can transform itself from an electric truck company into a legitimate autonomy and artificial intelligence powerhouse.
Tomorrow Everything Changes (Maybe)
On Thursday, Rivian is throwing open the doors for its very first “Autonomy and AI Day.” Think Tesla’s famous AI Day events, but focused almost entirely on vehicles rather than humanoid robots. CEO RJ Scaringe has promised deep dives into computing hardware, sensor suites, neural network architecture, and something he keeps calling the “data flywheel.”
In plain English: Rivian wants you to believe it’s building the brains that will one day let its trucks and SUVs drive themselves better than anyone else—except maybe Tesla and Waymo. And crucially, it wants to convince Wall Street that this software can eventually be licensed to other car companies, turning today’s billion-dollar losses into tomorrow’s high-margin recurring revenue.
The Harsh Reality Rivian Can’t Escape
Let’s not sugarcoat it. The core business is struggling. Through the first nine months of 2025, Rivian deliveries are down 14% versus last year. The company has cut costs aggressively (layoffs, supplier squeezes, you name it), but it’s still burning roughly a billion dollars a quarter. Even the blockbuster $5.8 billion joint venture with Volkswagen only buys time—it doesn’t solve the fundamental problem that building and selling $70,000-$90,000 electric trucks at scale is brutally capital-intensive when consumer demand is softening.
Every pure-play EV company is feeling the same pinch right now. But Rivian has less cash runway than Tesla and none of the brand halo that lets Lucid or Fisker keep raising money on dreams alone. Something had to change.
From “Adventure Vehicles” to AI Company
RJ Scaringe has always described Rivian as a technology company that happens to make trucks. Early investors bought the story about vertical integration, cloud-connected fleets, and over-the-air updates. But until very recently, the actual software features lagged years behind Tesla and even legacy players like Ford and GM.
That’s changing fast. The second-generation R1T and R1S (the refreshed models that started delivering this year) switched to a new zonal architecture with far fewer ECUs and dramatically more computing power. Translation: the trucks finally have the raw horsepower to run serious autonomy stacks in-house instead of outsourcing to Mobileye or Nvidia black boxes.
“Over the longer term, we believe what will differentiate Rivian’s autonomous capabilities will be our end-to-end AI-centric approach.”
– RJ Scaringe, Q3 2025 earnings call
That quote sounds suspiciously like something Elon Musk would say. And that’s exactly the point.
What Wall Street Actually Expects Tomorrow
Analysts have been circulating preview notes all week. The consensus seems to be:
- Updated timeline for “hands-free on any road” (currently limited to highways)
- First real look at the Gen2 perception stack running on the new Rivian Autonomy Platform
- Details on training cluster size and miles of data collected
- Whether Rivian is sticking with pure vision or quietly adding lidar for R2/R3 vehicles
- Any hints about licensing the stack to third parties (the real margin unicorn)
Perhaps the most interesting rumor floating around is that Rivian will show Level 3 capability on closed courses—meaning the vehicle can legally take full control under specific conditions, with the driver allowed to look away. Only Mercedes currently offers anything like that in the U.S., and it’s geofenced to tiny areas of California and Nevada.
The Lidar Question Nobody Wants to Answer
For years Rivian proudly declared itself “vision-first” like Tesla. But Scaringe has started talking positively about lidar in recent interviews, saying the company is “open to the right sensor suite for the right application.” Translation: the upcoming cheaper R2 SUV (launching first half 2026) might ship with roof-mounted lidar pods.
That would be a major philosophical shift—and a pragmatic one. Pure vision works great when you have 500,000+ vehicles collecting data every day (hi Tesla). Rivian has maybe 120,000 vehicles on road total. Adding lidar could let them leapfrog years of vision-only training data needs.
Can Software Actually Save the Company?
Here’s the bull case in a nutshell: if Rivian can build a world-class autonomy stack, three amazing things happen:
- Existing owners pay for subscription ADAS upgrades (like Tesla’s FSD)
- Future vehicles command premium pricing
- Most importantly, the stack gets licensed to legacy OEMs desperate for credible Level 3 solutions
That third point is the holy grail. Imagine Ford or Stellantis paying Rivian $5,000–$10,000 per vehicle for autonomy software instead of the other way around. Gross margins flip from negative to 70%+ overnight. Suddenly all those billions in losses look like R&D investment instead of money down the drain.
The bear case is simpler: Tesla has a five-year head start, Waymo has a ten-year head start, and the capital markets may not fund another multi-year moonshot when cash burn is already this high. Morgan Stanley literally downgraded Rivian this week citing exactly that fear.
Personal Take: This Is Make-or-Break
I’ve followed Rivian since the Amazon order days. The trucks are legitimately fantastic—probably the best pure EVs you can buy that aren’t a Tesla. But fantastic hardware only gets you so far when your market cap is $17 billion and you’re losing $5 billion a year.
Tomorrow isn’t just another investor day. It’s Rivian trying to rewrite its entire narrative from “premium EV also-ran” to “the next great American autonomy company.” If they show convincing progress—real rides, real data, real timelines—I think the stock could have another 50-100% upside even from here. If it feels like vaporware 2.0, the downside risk is severe.
Either way, the EV industry is watching. Because if little Rivian can pull this off, it proves the pure-play model isn’t dead. It just needs to evolve beyond selling metal boxes with batteries.
See you on the other side of AI Day.
