Taking Tesla's FSD (Supervised) Out for a Spin

Like many a Tesla-owner, I’ve spent the last month greedily taking advantage of Tesla’s Full-Self-Driving (Supervised) courtesy of their month-long free trial offer. When I bought my Tesla, I didn’t even consider purchasing the FSD option. After all, it cost almost half of what I paid for the car! But as a technology enthusiast, the chance to use it for free was just too cool to pass up.

So how is it? Does it work? What are the drawbacks? Did I kill anyone? Is my Tesla still unscratched? Will it succeed? Would I buy it and how much would I pay? I had all these questions – and I have answers now – or at least opinions – to share.

Let’s start with the biggest question: does it work? Yeah – it does. It’s pretty great actually. My wife and I have used it a number of times and it has handled almost every single driving situation with aplomb. It’s handled the narrow, winding, hilly road to my house (with plenty of blind corners, Marin mountain-bikers, and occasional deer) safely and well. Even when confronted by difficult and sudden situations (narrow road, pedestrian, car coming around blind turn in the other direction), it’s made reasonable decisions. It does a great job in normal traffic. It detects almost every possible threat and danger. And it mostly drives the way a person does – without some of the maddening quirks that afflict robotaxis.

Did it stumble a few times? Yes, it did. The biggest problem was at a broken stoplight where it wouldn’t take its appropriate turn at what amounted to an impromptu 4-way stop. It sometimes disengages when you don’t want it to. It drove 5 miles an hour on a private school road (posted at 5 miles an hour) – which was maddening but understandable. Worse, it didn’t recognize when it was off the private road, and it stayed at 5 miles per hour for about a mile too long. Obviously, if you’re not “testing” FSD, you can just kick it in gear in situations like this. It made a few unnecessary lane changes, which still puzzles me – I couldn’t figure out what triggered that. ?It missed one or two speed limit signs (what driver doesn’t). Once, it braked pretty hard at a yellow-light that most drivers would have gone through (I would have for sure).

On the whole, it’s an amazing piece of technology. It works well enough to use and use with pleasure in a wide variety of driving situations. Given Tesla’s troubled history with FSD and mostly terrible press, I was impressed. If Ford released its equivalent, pundits would be down on their knees kissing the tires.

So, what are the drawbacks? Well, let’s start with that (Supervised). The truth is that for most of us, the difference between driving and supervision is often infinitesimally small. If I have to pay attention, I might as well drive. And you do have to pay attention. And yes, as per the cartoon, I did have to remind my wife (who does most of the driving when we're together) a couple of times that she should enter a destination. Blame Tesla!

For most around town driving, we didn’t find it much of an advantage. Maybe not an advantage at all. Letting FSD do the work is very cool at first. But once you’re used to it, I’m not sure driving isn’t better than supervising. The (Supervised) also means that the people who might benefit most from full self-driving don’t benefit at all. My mom (who no longer drives) can’t use it though she can hop in a Waymo. That’s a shame, but I don’t see it changing anytime soon. On the other hand, my wife’s father still drives but might be better-off and safer as a supervisor.

?

There are a couple of use-cases where FSD(S) would be pretty great. On long road trips – especially on highways that are moderately or heavily trafficked but flowing, it would be awesome and would take almost all the work out of driving. It’s much superior to cruise-control. My wife gave a battle-cry of delight when the FSD pulled out from behind two slow-moving cars into the adjacent lane (on a non-highway road with 2 lanes in our direction) to pass. We thought it was just going to meekly follow along a few miles under the speed limit. No so, my friend.

Ditto if I had a daily commute with a lot of start-and-stop driving in heavy traffic. Stop-and-start driving is brutal and not only would FSD(S) take away a lot of the stress, it would probably make it safer since humans tend to be very bad at paying stop-and-start attention.

Did we kill anyone? Not even close. And no, no scratches either. There were very few intervention experiences and in all of them, it seemed like the FSD opted for the most conservative (though not always ideal) course. Drive long enough and you’ll have some close calls. Based on our limited experience, I’d say FSD(S) handles normal driving better than most drivers but is prone to over-conservatism or indecision in some unexpected ones.

Will it succeed? Unsurprisingly, it probably depends on your definition of success. At least in its (Supervised) form, I don’t think many people will pay the current $100/month price unless they have one of my compelling use-cases. But hey, it’s not like there aren’t plenty of commuters out every morning enduring stop-and-start driving. If I was still working for EY and going into the SF office three or four times a week, I’d at least think about it.? On the other hand, given how I drive, that price point is way too high for me and probably most drivers. Despite what I would describe as a very positive experience with it, I wouldn’t even consider subscribing.

So how much would I pay? Given my current driving habits, I think I’d probably give it a go in the $7-$12/month range. About what I pay for a mediocre streaming service or Apple Music. That’s not nothing, but it’s still a lot less than $100/month. I don’t think activation involves any costs for Tesla, so there may be no reason they can’t set whatever price maximizes revenue. My guess is that the optimum is either going to live fairly high (say $50/month) and appeal primarily to commuters and heavy drivers, or it’s going to need to swing all the way down into my range to become attractive to a much larger segment. ?I have no idea which point would be more profitable. That's what market research is for.

?

If you own a Tesla, don’t already have FSD(S), and have the free-trial offer, give it a try. You might be pleasantly surprised. For us, it was an interesting experience in a good way, and I’d love to hear how your trial goes!