Thanks for the well thoguht out reply, appreciate how you weave it into the history of automation.
In retrospect I suppose I should change my stance from being “near worthless” to being a nice option to have If you can afford it. I get what you are saying about the development Of FSD as a long continuous line, but from a business / product marketing perspective the moment it becomes unsupervised and does not require a driver is the moment it becomes an entirely different product. It’s no longer just a driver assistance feature, it’s a personal driver.
Some analogies that maybe highlight the product differences between unsupervised FSD vs supervised FSD:
A personal gardener vs some nice high end gardening tools.
A personal chef vs some nice high end kitchen equipment
A tailor vs a nice sewing machine
A security guard vs an alarm system
A builder vs some nice DIY equipment
A cleaner vs some powerful cleaning equipment
The former does the work for you without your presence or attention, the latter requires your presence, attention & some effort.
The value of a “nice to have option“ that is supervision required FSD to tesla is maybe tens of billions in market cap. The value of an approved unsupervised FSD product that requires no human presence is somewhere in the hundreds of billions to trillions (depending on whether other vendors also offer similar)
Hmm...each of those analogies appear to be presenting a human, and assuming the human does it all for you, including the decision making, and handles all possible situations, even those you don't encounter. But wouldn't:
1. A personal gardener still need your input from time to time ("That tree is leaning more and more, are you ready for it to be removed and lose the kids treehouse?"). Conversely, for many other homeowners with just grass in the front and back yard, the evolution from manual-mower to powered-mower to riding-mower to robotic-mower (already available), with automatic sprinklers, the high-end gardening tools of robotic mower and automatic sprinklers may give them the same experience for their needs.
2. A personal chef is fantastic if one wants home-cooked complex meals constantly but can still screw up ("Trying a new recipe for the chicken, hope you don't get food poisoning"). Whereas for someone who predominantly dines out / has food delivered and eats only very basics at home (at sporatic times of the day and night), a Keurig Eccellenza Momentum and a fridge to keep the creamer and some milk cold is more useful than telling the chef to make a bowl of cereal at 3am. I know some people who have literally not cooked a meal in incredibly long times.
<...>
As I look at the list, it occurs to me that in most cases, the first item would likely need the second item (tailor needing a sewing machine, guard still needing all the windows / doors etc to be alarmed, cleaner still needing a vacuum, etc). And, when something goes wrong with the second item on that list, you-the-owner will still need to get involved (it will be you, not the security guard, dealing with the police report and insurance company, etc). The greater the complexity of what is needed, the greater the need for the first item. That's actually kind of my point, in reverse...the more things can be simplified, the less the first item is needed.
FSD for someone who only goes to places near their newly built suburban neighborhood in an upscale area of the US, where all streets are nice and wide and newly marked and parking is always immediately available and they don't mind walking a block or so from parking to destination, etc, will likely get 100% solved
for them and their specific use case first. While they are celebrating, others who have more complex transportation needs will still be waiting because it won't be solved
for them and their specific use cases.
If one sets the goal of "FSD isn't done until it doesn't require anyone's attention anywhere in any situation worldwide", then the timeline may be in the many-decades magnitude. However, long before then, it will be solved to not require anyone's attention the vast majority of time.
Consider it this way: Assuming you lived in the most FSD-compatible house in the most FSD-compatible neighborhood possible, and had the most FSD-compatible travel needs, etc etc etc, how many years do you think it will take for FSD to handle 5% of your rides independently without you paying attention? 50%? 100%? Now assume AAA (or your auto manufacturer) can not only come to you to fix your unexpected flat tire, they can also remote-control your car (if you give them permission) to get it past that unexpected weird thing that happens once a year that FSD wasn't able to handle...what does that timeline look like?
TLDR:
The value of a “nice to have option“ that is supervision required FSD to tesla is maybe tens of billions in market cap. The value of an approved unsupervised FSD product that requires no human presence is somewhere in the hundreds of billions to trillions (depending on whether other vendors also offer similar)
I think you are underestimating the value of "I have to supervise but rarely have to do anything", but even with your framing, there are the in-between values of "It can take grandma to her weekly knitting circle down the street for me, so now I only need to drive her to the 17 doctor's appointments this week"..."It can take my 12-year-old to their school 2 miles away and pick them up afterwards while I'm at work, and the twice-a-year the car gets stuck I get an alert and approve remote-control driving by AAA or Tesla"..."Once I'm on the freeway I can stop paying attention for the next 200 miles until it gets to within 5 miles of my exit". Even just looking at the "unsupervised FSD product" scenarios, it will *not* be a single day when everything works for everyone...it will be a gradual release of increasing scenarios being handled. And each one of those will add significant value long before any 100%-unsupervised-in-any-scenario product is anywhere close to reality.
