Tesla to upgrade all HW3 cameras and computers to HW4 - RJ, your move.

[Zorg]

My guess is that it'll never happen. Instead, discounts will be offered for a new car and that's that. The rest will get refunds through class action settlements

Sponsored

 

[ksumnole]

Well now a precedent has been set.

Rivian? G1 upgrades?

Haha

Good troll OP. Either you are living under a rock or think Rivian owners are.

 

[Motoarzon]

My guess is that it'll never happen. Instead, discounts will be offered for a new car and that's that. The rest will get refunds through class action settlements

There are approx 4Million cars with HW3 hardware and about 600k customers who who have HW3 "and" have purchased FSD outright. It's these 600k that Tesla owes a solution. The other 3.4M customers Tesla owes nothing. They can't have compensation for a product they "did not" purchase.
-Musk said on the earnings call they will offer those with HW3 a discount to a new vehicle. They would offer and option for a either a full refund for their FSD purchase and later down the road when/if you decide you want FSD you just do the subscription model, or simply transfer your FSD over to the new vehicle which Tesla already done incentives before.
This will take care of most of these 600k as a lot of those cars are getting old and due for the newest latest and greatest anyway.
-For everyone else they would simply lock HW3 cars with no further significant improvements and leave FSD "supervised" at the capabilities it is today (which is still pretty darned good!! and most people are fine with and don;t car for FSD). Perhaps lower the monthly subscription to $50 to reflect it's current capability as opposed to the HW4 cars getting better and better and eventually full "unsupervised"
-There is still a huge business case either way to upgrade all HW3 vehicles. $2k of hardware and effectively $1k labor ($3k) is peanuts compared to a customer paying $100/mth forever (a lot more $$$ as the software becomes fully "unsupervised"). Plus, when unspervised and used on the robotaxi network they will easily make at least $36k per year in revenue. 10-20 short rides per 24hrs is at least $100 x 365 days =$36k each and every year thereafter.

 

[SolartoEV]

-There is still a huge business case either way to upgrade all HW3 vehicles. $2k of hardware and effectively $1k labor ($3k) is peanuts compared to a customer paying $100/mth forever (a lot more $$$ as the software becomes fully "unsupervised"). Plus, when unspervised and used on the robotaxi network they will easily make at least $36k per year in revenue. 10-20 short rides per 24hrs is at least $100 x 365 days =$36k each and every year thereafter.

Why limit it to at least $36,000? Maybe it will be 36 Million.

I love speculative pure bullshit math!

 

[Zorg]

Obviously if one didn't purchase FSD, one isn't owed anything. Figured this much was obvious.

More importantly, 600k FSD purchases at say $8k a pop is a potential max refund liability of $4.8B. Put that way, this may be a great incentive to upgrade those owners.

 

[BigSkies]

The key difference is that Tesla marketed their cars as having all of the hardware features necessary for truly autonomous operation, and even made claims they'd be appreciating assets because of it.

If Rivian marketed their Gen 1 vehicles this way, then a similar expectation would be reasonable.

I don't remember any Gen 1 claims around this. I think there were some Gen 2 claims about the system improving over time, but I don't remember them being specific. I don't really care to look though.

 

[Ralph]

My guess is that it'll never happen. Instead, discounts will be offered for a new car and that's that. The rest will get refunds through class action settlements

You might be right; but consider:

Any statements he has made relative to building min-factories to retrofit tech required to fulfill promises made are potentially being driven by multiple factors:

• lawsuits / class actions (his statements serve to drive discussion and bargaining positions)
• his pay package with incentive for more subscribers
• mini factories useful for some other unstated purpose
• Who knows?

It's not hard to envision a scenario where the customer can retrofit at some "discounted" rate. The customer gets access for the "reduced retail price", but part of the retrofit deal requires the customer to subscribe for X period of time.

An amount of time that just happens to fulfill a requirement in his pay package to increase subcribers.

 

[Donald Stanfield]

This retrofit is going to be implemented the same time the extra batteries are going to be installed on the cybertruck and the same year the roadster with super secret rocket boosters will come out.

 

[Motoarzon]

Why limit it to at least $36,000? Maybe it will be 36 Million.

I love speculative pure bullshit math!

Limit, who said limit? Speculation? LOL. Simple "estimated" math. Most 10-20 minute uber rides are $15-$20 and sometimes way, way more depending the market and time of day.
A Robotaxi likely does it for 50% less which is doable as your fuel is cheaper and you don't have to give 30% to pay a driver.
If a robotaxi does even 2 rides per hour for $10 thats $20/hr. Lets say it does ride share for 20hrs per day (4hrs for charging and/or cleaning). Thats still $400 per day. Thats $146k/yr revenue. Net profits would likely be at least $36k.
Again, people have no clue the gravity of robotaxi.

 

[Zoidz]

Well now a precedent has been set.

Rivian? G1 upgrades?

What precedent?

Oh, Elon making unsubstantiated claims that never come true, got it.

 

[CharonPDX]

Tesla promised in late 2016 that "all vehicles shipping now have all the hardware necessary for full self driving" - and sold those vehicles with that commitment.

Rivian has never sold their vehicles as "can do unsupervised level 4/5 self-driving with the hardware in it today."

Tesla *already* set the precedent, when they upgraded HW2 vehicles to HW3, because HW2 wasn't advanced enough.

 

[SANZC02]

Limit, who said limit? Speculation? LOL. Simple "estimated" math. Most 10-20 minute uber rides are $15-$20 and sometimes way, way more depending the market and time of day.
A Robotaxi likely does it for 50% less which is doable as your fuel is cheaper and you don't have to give 30% to pay a driver.
If a robotaxi does even 2 rides per hour for $10 thats $20/hr. Lets say it does ride share for 20hrs per day (4hrs for charging and/or cleaning). Thats still $400 per day. Thats $146k/yr revenue. Net profits would likely be at least $36k.
Again, people have no clue the gravity of robotaxi.

You are not factoring in competition. Your estimate has you getting 40 rides per day. Uber averages ~30 million rides a day. Not counting existing drivers or competing vehicles, if Tesla sales 5 million vehicles that start ride sharing based on current averages that is only 6 rides a day per Tesla in service. Add in other drivers/vehicles it would likely be less than that.

Now being generous and say you get your 6 daily average at $10 per you are best case scenarios 21.9k gross annually. Then factor in expenses, start with electricity, taxes, tolls, insurance. Now factor in wear and tear on a vehicle not built with the durability of a commercial vehicle. Then think about the companies building fleets for this, Wemo, Tesla, Pony,ai, zook, etc.

Not saying you can’t make a little money trying this but be realistic in your estimates. Here is a recent article where real drivers share their income data for reference.

 

[Riviot]

If Rivian marketed their Gen 1 vehicles this way, then a similar expectation would be reasonable.

I don't remember any Gen 1 claims around this. I think there were some Gen 2 claims about the system improving over time, but I don't remember them being specific. I don't really care to look though.

Rivian has never sold their vehicles as "can do unsupervised level 4/5 self-driving with the hardware in it today."

Video from 2019, explained as featured in 2020 tech.

 

[Zoidz]

You are not factoring in competition. Your estimate has you getting 40 rides per day. Uber averages ~30 million rides a day. Not counting existing drivers or competing vehicles, if Tesla sales 5 million vehicles that start ride sharing based on current averages that is only 6 rides a day per Tesla in service. Add in other drivers/vehicles it would likely be less than that.

Now being generous and say you get your 6 daily average at $10 per you are best case scenarios 21.9k gross annually. Then factor in expenses, start with electricity, taxes, tolls, insurance. Now factor in wear and tear on a vehicle not built with the durability of a commercial vehicle. Then think about the companies building fleets for this, Wemo, Tesla, Pony,ai, zook, etc.

Not saying you can’t make a little money trying this but be realistic in your estimates. Here is a recent article where real drivers share their income data for reference.

How dare you challenge the Robotaxi narrative with actual facts!

 

[Zorg]

Robotaxi, if it actually works as a level 4, is chasing the Uber/Lyft market. Except it will make sense mostly for fleets. The narrative of individual buyers renting out their Robotaxi seems far fetched to me.

Sponsored