Tech enthusiast 🚀 | Tesla & SpaceX Fan | Inspiring Automation | Enjoys political discourse | Championing business innovation | Theme Parks | Cruise Addict
The future is here! 🤯
My Model Y Tesla just completely navigated a go kart track on FSD! It did this on FSD 12.5.4! Cannot wait to try it again with V12.6 on Hurry Mode!
I feel like this is going to make @cybertruck jealous.
PS. It worked with Actually Smart Summon also. 😎
I just got another unsupervised robotaxi.
Tried to get in the driver’s seat.
Tried hitting the pull-over button.
Tried everything.
Unleash the robotaxis.
Well done team
As one of the most senior ADAS experts in my company, and as people are slowly starting to realize that FSD is real (after denying it all along), I often get asked: how can we compete against FSD?
My answer invariably shocks them: we can't.
@FredaDuan Another major question are the OEMs ability to integrate and engineer this AI solution when they can barely do an infotainment system that rivals an Atari interface and even that they outsource 🙄
It is striking (though not surprising) how poorly legacy automakers have invested in their future. Technology is reshaping transportation at an accelerating pace, yet most incumbents are not playing to win. They are playing to delay.
Their first failure was cultural and structural: a broad rejection of the EV transition. By any rational measure, legacy OEMs could have simplified their product lines, collapsed complexity, lowered costs, and built more efficient vehicles. Instead, progress stalled because dealers (the real customers of legacy auto) resisted change. EVs threaten the service revenue that underpins the dealer model, and the OEMs remain captive to it.
Those that did pursue EVs made a second, more damaging mistake: they refused to start from a clean sheet. Rather than building modern, software-defined vehicles, they bolted electric powertrains into architectures designed for ICE cars. The result was predictable— bloated wiring, fragmented ECUs, vendor-owned software, and vehicles fundamentally incapable of rapid iteration, cost compression, or true OTA evolution. They built electric cars, not modern machines.
Autonomy exposed this weakness completely. Legacy highway-assist systems were never designed to win; they were designed to check a feature box. With no data flywheel, no vertically integrated stack, and no AI-first architecture, most OEMs outsourced the “brain” and accepted permanent dependency. Autonomy, for them, was a feature. For winners, it is the product.
The problem is not that legacy CEOs are blind. Most understand exactly what is happening. But they are trapped, by dealer economics, union constraints, internal incentives, stranded assets, and balance sheets that cannot survive a true reset. So they allocate capital defensively, mistaking spend for progress and compliance for strategy.
Where does this leave them? Not suddenly extinct, but structurally behind, with futures defined by dependence and protection rather than innovation and control. These are survival paths, not winning ones.
The auto industry is being rewritten in software, manufacturing systems, and AI. Legacy auto largely chose not to compete there. The consequences are now unavoidable.