What do sensors know?

person driving car during night time
Photo by Gabe Pierce on Unsplash

China’s new rule: if your car uses sensors to map the environment, you must apply for government permit (link). Car sensors are now so good and ubiquitous they’re viewed as security threats. ‘The statement is a clarification of China‘s surveying and mapping law and reflects regulators’ efforts to prevent highly-detailed visual data collected by smart cars falling into the hands of hostile foreign actors.’ Related: A new sign shows that Tesla cars cannot enter government agencies and state-owned enterprises in China (Trucks VC)

Chinese decisions on car sensors are a testimonial that, as a nation, they genuinely know the capabilities of sensors and connected sensors as data and intelligence sources! As the saying goes, it takes a thief to know another

Did you notice: That we have not heard a peep from Elon Musk about why Chinese regulators haven’t approved  Tesla’s FSD for use on local roads? That man is always yapping about US politicians, loudly and crassly — because, you know, democracy and shit! 

September 22, 2022. San Francisco

It is not often you get to do something you have never done before. That is an argument made by a close friend, who has been urging me to take on everything uncomfortable and difficult during what is a personal transition. He often reminds me that being uncomfortable is what has kept me moving forward, evolving, and perhaps improving. So, after some misgivings, internal anxiety, and lots of fears, I embarked on a new adventure today.

I took my first driving lesson today. (In a somewhat ironic turn of events, someone else took their first driverless robotaxi ride in San Francisco.)

Driving hasn’t been a priority for me. I grew up in Delhi and learned how to ride a scooter and motorcycle. It was enough to get around, and more importantly, that was all my family could afford at that time. My parents taught me to live within our means. It is a lesson that isn’t easily forgotten even after all these years. 

Even though I had friends with cars who would have readily obliged, I didn’t feel comfortable asking them to teach me. Nevertheless, I could remember roads and quickly navigate, which landed me some navigator gigs in car rallies. I still never drove myself. Later, when I moved to New York, I found no need to drive. MTA was a godsend. Even in San Francisco, I found myself taking public transportation and later taxicabs. Uber solved many of my mobility problems.

Owning a car, the carbon footprint of a car, and just dealing with the bureaucratic complexity of ownership made me not bother with driving. Even today, I would rather be part of a collaborative transportation system,

However, as my passion for photography has increased, I have found myself constrained by the limitations of on-demand transportation. I have often missed the morning light because I couldn’t get to the desired spot in time. My friend, who encouraged me to take on the challenge, reminded me that I could drive to destinations that opened up more photography opportunities. 

I honestly (and perhaps naively believed) in technology to bring driverless self-driving cars to the masses and that I would never have to learn to drive. Let the machines do what is clearly a highly complex task. 

Despite Elon Musk & Co.’s best efforts, the recent launch of Cruise’s driverless robotaxis, and all the hype, it might be some years before a driverless future becomes a reality. That doesn’t mean that I doubt the future — the chips, the lasers, the software, the networks, and everything else is improving exponentially by the day. 

However, should I wait or take matters into my own hands? The answer was pretty obvious. And for the first time in my life, I sat behind a steering wheel, and in the company of a driving instructor, I drove for about three miles on roads with actual traffic. Most of the session was about learning how to find the delicate balance between accelerating and decreasing speed. The accelerator and the brakes are two sides of the same coin, really, as long as you can find the right balance. I suppose in time, and with enough experience, you start to find harmony. It will perhaps be an excellent way to measure how far I have come along on this journey. 

So how were my first two hours behind the wheel? Nerve-wracking, to say the least, even though my instructor thought I was relatively calm. I got back home and exhaled — drenched in sweat and relieved that I had passed the first hurdle. 

As I sat down and replayed the two hours in my mind, I could start to see my specific challenges and driving challenges in general. I gripped the steering wheel too hard, and that made sometimes keeping in the lane a bit tough, as my tension made the car twitch a bit. I need to learn to breathe easier behind the wheel, and perhaps they will allow me to hold the steering easier. It is what I need to start to develop trust in the machine itself. 

More importantly, the whole process made me confront myself — I like to have everything controlled and managed, where I have enough time to deliberate and make decisions. That is part of the analytical approach that has become part of my daily rigor. Driving is nothing like that — it is almost the exact opposite. You need the staid approach of a grownup with the reflexes of a teenager in order to be an alert and efficient driver. 

Even though it was my first attempt at driving, it was evident that driving is not about being the perfect driver. Instead, driving is about being intelligent and agile enough to deal with the imperfection of the world around you and assuming the human ability to do irrational things. 

You have to pay so much attention to multiple data streams entering your brain when driving on the road. It is a bit overwhelming, and more than anything, makes you appreciate how vital self-driving technology will be for mobility. It also made me appreciate the complexity that all these automation systems will have to deal with, reinforcing the importance of data collected by these self-driving companies. You also start to realize the vital edge Tesla has over its rivals when to all the training data it has collected and algorithms it has crafted as a result. 

While self-driving and its proponents come under intense criticism — and why not, they do overhype things too much — we can’t but be excited about the possibilities. I was not fortunate enough to be around at the birth of personal computing, but I did see the rise of commercial internet from a courtside vantage point. The same was true for the rise of mobile computing. 

Self-driving and automation are new kinds of computing environments, and their journey is exciting for me as a technology enthusiast. I have been following the emergence of new kinds of chips and technologies. As an investor, I have invested in Veniam, which enables the networks that power this automation. 

And as a student driver, I will do my best to get my driver’s license. I am not sure if I will even buy a car. All that is in the future, as is the time, when I can say: hey Tesla, drive me to Hawk Hill!

November 13, 2021. San Francisco.

A few weeks ago, Jonathan Stern got in touch, wondering if I would like to participate in a new website he and his best friend, Carter Duncan were creating. “In the spirit of the Federalist Papers, my best friend from Duke and I have built a website called Pairagraph for written dialogue between distinguished individuals,” he wrote. The thinking behind the project was to “bring world leaders across government, industry, academia, religion, and the arts into a conversation about great events and great ideas, and perhaps, to revive the Republic of Letters.” 

University of Berkeley economist Brad DeLong is interested in discussing whether the U.S. is in decline. Is it true, as pessimists proclaim, that America has plateaued? Or that we are mired and trapped in civilizational languor? Is our culture exhausted? As someone who loves Prof. DeLong and his writing, it was such a privilege to have a dialogue with him. 

While the good professor is feeling discouraged about the American prospects, I am a long-term believer. I posted my first response today on Pairagraph! My original draft was over a thousand words, but I can’t go beyond 600 words. For now, here is my abbreviated take on why I don’t think America is in decline despite the recent events. Not yet!   

Tesla’s SolarCity gamble has gone wrong

“If he hadn’t bailed out SolarCity, his whole debt-laden empire might have cracked. Yet without the bailout, Tesla would be far more healthy….In the second quarter of this year, SolarCity installed only 29 megawatts of solar panels—far below the 10,000 megawatts in annual installations that Musk had promised.”

Vanity Fair

Bethany Mclean who made her name writing about Enron long before others is explaining the challenge faced by Tesla and Elon Musk due to the 2016 SolarCity acquisition. The much ballyhooed Solar Roof is a flop, and the whole thing seems to be coming apart at seams. Worth a read from a reporter, who has a habit of finding big stories before others.

Read article on Bethany Mclean

Details on Tesla’s new big beefy chips

Each Tesla computer has two AI chips, a redundant design for better safety, Venkataramanan said. There’s redundancy in the chips’ power supplies and data input feeds, too. Even the car’s cameras are on two separate power supplies to guard against failures. ….Each Tesla AI chip runs at 2GHz and performs 36 trillion operations per second. That performance is possible because Tesla optimized the chips for self-driving cars and dropped anything more general purpose…..For example, the chip handles data recorded as 8-bit integers instead of the 16-bit floating-point numbers more common in AI tasks but that require more power to process. For another, it’s got an extremely limited set of instructions it can process. And it’s got a gargantuan 32 megabytes of high-speed SRAM memory on the chip, which means it doesn’t have to wait around while fetching data from much slower conventional DRAM memory.

Apple has taught Silicon Valley the importance of owning its own chip destiny and now pretty much everyone willing to push the technical edge is building their own chips, for more vertical integration in their designs. Tesla gave a glimpse into how it can keep ahead of its deep pocketed rivals. At the Hot Chips conference, Tesla showed details on its newest innovation. The chip took 14 months to design and Samsung is going to make the processor. It is in newer Tesla cars. It is 21-times faster than the Nvidia chip they were using and about 80 percent of the cost.

Also: live blog from HotChips conference that has more details.

Read article on C/Net

Waiting for (Autonomous) Godot

Rodney Brooks, founder of Rethink Robotics and a robotics guru is skeptical of the current claims around AGI and autonomous cars hitting the road anytime soon, much like another technology luminary, Kai-Fu Lee.

If we were to have AGI, Artificial General Intelligence, with human level capabilities, then certainly it ought to be able to drive a car, just like a person, if not better. Now a self driving car does not need to have general human level intelligence, but a self-driving car is certainly a lower bound on human level intelligence. Urmson, a strong proponent of self driving cars says 30 to 50 years.

I guess the biological genes have to live to a 100 to see their arrival.  But, do read Rodney’s essay, to understand his skepticism around the near term claims around autonomy.   That said, I think we need to lower our expectations. Autonomous versus smarter electric cars should be our expectation. Unfortunately, that kind of simple thinking doesn’t prop up the stock.

Read article on Rodney Brooks

What Did Tesla Do Now?

In many ways, Tesla — Elon Musk’s lightning rod of a car company — is the perfect allegory for modern Silicon Valley. The ongoing psychodrama of personalities drowns out the amazing technical achievements that are happening all around us. While the latter is what will ultimately determine our future, it’s the soap opera most of us can’t turn away from.

As usual, this has been a real “Dr. Jekyll and Mr. Hyde” kind of week for Tesla. It had a disastrous earnings report card, and Elon keeps creating all the wrong sorts of headlines. But in the middle of this maelstrom, the company announced a new chip that is going to eventually become the brain for their electric car. This chip is not just any chip — it will be able to make sense of a growing number of sensors that allow the car to become better and better at assisted (if not fully automated) driving. Continue reading “What Did Tesla Do Now?”

There is something about (Elon) Musk

The last month in the life of Tesla and its mercurial chief executive can best be described as a bungee jump from a plane. No surprise, that we have a bull market in opinions about Elon Musk, his behavior, and his actions.

I don’t own Tesla shares. I have no desire to own them. Just as I don’t want to short them. Hell, I don’t even drive.

But you know what I dream of? Owning a self-driving electric car before I leave the planet. And the odds of that happening are way more likely if Tesla and Elon are around.  Continue reading “There is something about (Elon) Musk”