this week i'm thinking about – #10 – remote-controlled cars
Thanks for sticking with me for 10 full issues! Feedback and suggestions are very welcome.
I’m writing a weekly blurb about something I learned that’s broadly tech-related. If you have thoughts, I'd love to hear from you, and I'll paraphrase the best responses in the next newsletter. My goal is to start conversations with people thinking about similar topics through different lenses. If you know anyone who would be interested in the discussion, please do forward this along or have them subscribe here.
What I learned about: remote-controlled cars 🎮🚘
The standard narrative around self-driving cars is that the labor efficiency gains and freed-up driver time will have ripple effects across the economy. But as the list of edge cases proves to be longer and harder to solve than originally anticipated, some creative technologists have come up with short-term solutions to circumvent the remaining level 5 CV challenges. Specifically, I’ve found remote-control error handling to be one of the more creative solutions. Phantom Auto, for example, provides remote-control safety drivers as a service. And some full-stack players like Zoox are building this competency in-house.
But when you expand the scope of remote-control drivers to the entire driverless car problem, you can actually get nearly identical benefits to level 5 cars with much lower R&D overhead: full error handling coverage, human-level accuracy, time saving from being “driverless”, and labor savings by using much cheaper outsourced labor.
Full remote-control driving is best tested in private use cases like construction sites, where a network outage won’t cause a traffic jam or accident. For a broader consumer use case, you’d need a low-latency wireless network with perfect coverage, which requires massive country-level commitment to investment in infrastructure. In other words, you probably won’t see it first in the USA 😊 I’d keep an eye on countries like China, Singapore, and Germany.
What I’m wondering about: European unfair advantages 🇪🇺
I am headed to Europe next month to see how the major tech hubs are evolving. What are technology problems that Europe is uniquely suited to solve? Some potential candidates: gaming (Supercell, King Digital), fintech (Revolut, N26, Adyen), travel (Trivago, Skyscanner).
Top replies from last edition (virtual beings):
Gabriel pointed out that AI-based facial reenactment is just about here (source), and text-to-speech is already fairly robust (source). Still lots of technological developments remaining for virtual friends, but many core elements are there.
Sacha shared The Waldo Moment, a Black Mirror episode in which an AI runs for public office. It shows one possible version of a future with virtual beings.
In response to the Facebook Class article, Jungwon thinks it’s worth developing a more nuanced framework for determining when to use a “move fast and break things” mindset versus “extreme planning and testing” in product development. I suspect we’re moving towards an era where the latter dominates across most industries.
What I’m reading 📖