Pinned Tweet
If you own a HW3 car, Wednesday night is your night.
Tesla reports earnings Wednesday, and shareholders have been voting all week on which questions get asked on the call. One of the most backed questions on the list is the one you have been asking all year. When do HW3 owners get the upgrade?
You know the history. Tesla promised HW3 owners a path back in April, a discounted trade in or a computer and camera swap. Six months later there is still no price and no date, and v14 Lite is where your cars stop.
The voting closes Tuesday at 4 PM Eastern. Wednesday night Tesla either answers it in front of everyone, or skips it in front of everyone.
So, HW3 owners. What do you need to hear Wednesday? A real date? A price you can plan around? Or are you done waiting on soon?
If Tesla sold FSD outright again tomorrow, would you buy it?
At the old $8,000, here is the choice as it actually stands.
Buy it on HW3 and you are paying for the ceiling. Lite is where it ends. No v15 Lite announced, and Musk has said HW3 cannot do unsupervised.
Buy it on HW4 and you are paying for a car whose computer gets replaced in production. AI5 reaches volume in mid 2027. And the purchase stays with that car forever. Transfers ended March 31.
Keep paying monthly and the math is $99 a month against $8,000, which is 81 months, about 6.7 years, to break even. Cancel anytime. Nothing sunk.
Or jump ship to Rivian. Autonomy+ is $2,500 once or $49.99 a month, with point to point reaching first owners in December.
Four very different bets.
Which one are you making?
HW3 owners already know the trade up math changed this year. Transfers ended March 31, outright purchases ended February 14.
Here is what that actually means if you trade up. The FSD you bought stays with the old car. Owners paid 8,000 dollars at the last price, some 12,000 or 15,000 earlier, and none of that follows you to the new car. You cannot buy it again either. The new car puts you on 99 dollars a month, which is 1,188 dollars a year, every year, with no buyout at the end of it.
That is on top of the trade itself. The car you trade in takes its normal depreciation hit, and the FSD line in your math flips from an asset you owned to a subscription you rent.
Waiting is the other side of it, and it costs nothing upfront. Tesla promised owners who bought FSD a free computer and camera upgrade. But there is still no date, and Musk has floated skipping to the next generation computer around mid 2027. Meanwhile v15 is headed to HW4 cars with no Lite version announced for HW3, and unsupervised is targeted around the second quarter of 2027 on HW4 hardware.
And waiting looks worse outside the US. In Australia, HW3 owners paid for FSD years ago, some at 10,100 Australian dollars, and still cannot use it at all. FSD there runs on HW4 cars only, and Tesla Australia still gives no confirmed timing for HW3, not even Lite. Most American HW3 owners wait with Lite in hand. Australian HW3 owners wait with no FSD at all.
So the choice is not just Lite versus new. It is a free upgrade with no date against a new car plus a subscription that never becomes yours.
HW3 owners, if you priced a trade up since transfer ended, what gap did Tesla quote you?
HW3 owners waited more than a year on v12 for this.
When Lite arrived in the summer, it was real progress. Smoother driving, and features HW3 had never had.
It also came with owner reports of cars throwing red hands and refusing to drive until the computer cooled down.
Now v14.3 Lite is here. So where does yours still make you take over?
HW3 owners keep asking whether v14 Lite is the real thing or a consolation prize. Months into it, here is the honest answer. It is genuinely good, and it is not the same car HW4 owners drive.
What it does best is everyday driving. Merges, forks, unprotected lefts, roundabouts, pedestrians, traffic lights, cut ins. Smoother steering, fewer false slowdowns, steadier lane centering. It parks, unparks, reverses, and lets you pick where it drops you, with Speed Profiles always on. That is a real jump from the v12 branch HW3 sat on for so long, and it is distilled from the same HW4 v14 brain, not a separate old codebase.
Where it falls short is the hard stuff. Reactions trail the full version, and the gap shows in complex edge cases, not on an easy commute. HW3 misses Automatic Collision Evasion, the feature that steers around a crash while you are driving manually. It misses the Mad Max profile and the new Smart Summon. Older S and X cars on Intel chips miss even more, no start from park, no arrival options, no reversing. Early Lite builds even had a pulsing feel owners flagged. Tesla fixed that in later builds, but it showed how tight the fit is.
That tight fit is the whole story. HW3 has about one eighth the memory bandwidth of HW4, and 1.2 megapixel cameras against HW4 5 megapixel. The modern FSD model simply does not fit on HW3 without being shrunk. That is why Lite exists, and why Musk said on the Q1 call that HW3 cannot reach unsupervised, no matter how good the software gets. V15 is headed to HW4 cars, with no Lite version announced for HW3.
So can Tesla close the gap? For supervised daily driving, mostly yes, and Lite should keep improving as HW4 teaches it. For the full HW4 experience, no. Unsupervised on this hardware is off the table, and the feature gap will widen as v15 lands.
That leaves HW3 owners three real options. Keep driving Lite, which is good and still getting better. Wait for the retrofit Musk promised to owners who bought FSD, computer and cameras both, though timing and the exact hardware are still unannounced, and he has floated skipping to the next generation computer around mid 2027. Or trade in for an HW4 car, which is the only route to unsupervised and the Q2 2027 date Ashok gave on the last call.
So HW3 owners, what is your move? Stay on Lite, wait for the retrofit, or trade up?
Yesterday Rivian pitched the future. Today it bought a piece of it.
Rivian just acquired Third Dimension, a UK startup that turns real images and video into lifelike 3D worlds. Fifteen people, absorbed into Rivian's autonomy team. Price not disclosed.
Here is why Tesla owners should care. This is Rivian building an FSD rival in the open. Point to point driving this year, personal Level 4 cars in a couple years, R2 robotaxis on Uber in 2028. What was missing was the validation engine, the part that proves the system is safe to ship. That is what this buy is.
Self driving lives or dies on edge cases, the weird stuff that almost never happens. Rivian gets 400,000 driving sequences a day from its own cars. Third Dimension lets it replay real scenarios with moved cameras, shifted traffic, different light and weather, without waiting for those rare moments to happen on the road. That is exactly the muscle Tesla built its FSD lead on: data, multiplied by simulation.
Fifteen people will not outdata Tesla. But the direction is now unmistakable. Rivian is assembling a full FSD competitor, piece by piece, in the open.
So for HW3 owners still waiting on an upgrade answer, and Hardware 4 owners watching their Q2 2027 unsupervised date, the question from yesterday stands, sharper now. What moves you? A demo, a date, or a car in your driveway?
Wednesday night Tesla would not even answer the HW3 question.
Last night Rivian's CEO took a reporter for a ride through New York and laid out what comes next. Point to point driving, hands off and foot off, in New York this year. Demoed in San Francisco in December, then sent to Gen 2 R1s and every R2 over the air.
And then he went further. The same Level 4 kit goes into personal cars. Your own car dropping you at the airport. Picking someone up for you. Nobody in the driver seat at all, a couple years out. Plus a robotaxi with Uber in 2028.
So here is the question for Tesla owners.
HW3 owners, you got silence on the call Wednesday night. HW4 owners, you got a Q2 2027 date for unsupervised. Rivian is now pitching you a car that drives itself with nobody up front, at the same price as a Model Y Performance, on Tesla's own Superchargers.
If Rivian actually ships this, who switches?
The Scaringe ride: xitter.cf/rhensing/status/211340…
They had the question. 667,000 shares behind it. Last night, they skipped it.
Go read the call. The investor questions covered Cybercab scaling, Optimus training data, unsupervised timing, Cybercab for sale, robotaxi costs, Optimus costs, and mapping. The HW3 upgrade question is not in there. Not answered, just not asked.
And here is the part that stings. The question right above it on the list got an answer. Unsupervised FSD in customer cars: Ashok said somewhere around Q2 2027. So HW3 owners now have a date for the thing their cars cannot do, and still no answer on whether their cars get the upgrade to do it, free or otherwise.
About four million cars were sold on a promise their hardware cannot keep. Last night was the night to say what happens to them. The call ended at 6:30 Eastern without saying it.
HW3 owners, after last night, are you still waiting, or are you done?
Tesla owners paid for the robot bet this quarter. Revenue flat at 28.05 billion dollars, operating profit down 56 percent to 718 million, because spending jumped 39 percent to 4.78 billion. The same update puts the future lineup into production. Semi in volume production in Nevada. Optimus on high volume lines in Fremont. Cybercab rides in Austin since September 3, nearly 2 million unsupervised miles by quarter end, and the next cities still have no dates. And FSD is now on more than 75 percent of new deliveries in North America, though Tesla does not say how many buyers keep paying after a trial ends. So the car business funded all of it. What do owners get back first, a Robotaxi ride in a new city or FSD worth paying for every month?
Tesla put up its first Robotaxi safety report today, and Ashok Elluswamy launched it on X himself. 3.1 million driverless miles through Oct 18 on the page, and Ashok puts the total to date at 3.3 million. Tesla's numbers are nobody hurt, no airbags, no pedestrians or cyclists hit. That is up from 380,000 on the July call and 1 million in early September, and about 70 percent came since late August when the Cybercabs started scaling. The fine print matters. Remote operator trips count, monitored miles from before Jan 22 don't, and two disputed injury claims didn't make the cut. You can download every NHTSA crash filing from the page, which is new. Nine days from now Tesla answers NHTSA's certification audit, and NHTSA opened the school zone speeding petition Oct 9. So the question is what the Oct 30 filing says that this page doesn't.
Full report on Tesla's page, launched by Ashok today. tesla.com/robotaxi/safety xitter.cf/aelluswamy/status/2112…
We are launching the Tesla Robotaxi safety report today.
Till date, Robotaxi has driven a total of 3.3 million miles autonomously, and has had zero injuries nor airbag deployments. If the same number of miles were to be manually driven, we should expect roughly 9 injuries to the occupants or other pedestrians and 4 airbag deployments on average.
We will continue to transparently share our progress with the public using this page.
tesla.com/robotaxi/safety
Your 2017 Model X just got its first v14 build last week. Tonight is the night owners find out if there is a plan beyond Lite.
FSD v14.3 Lite reached the Intel Model S and X cars last week, the last HW3 group with no v14 branch at all. Nine year old cars, running a distilled version of the new stack. Yesterday, HW4 cars moved on to v14.3.13. Same week, same company, two different futures.
The HW3 upgrade question finished at number 7 of 840 on the shareholder list, with 667,000 shares behind it. That is nearly double where it sat two days ago, and it still fell a spot, because late whale votes moved in above it.
Here is the part nobody is talking about. The number 2 question on that list has 5 votes. It sits above yours because Say counts shares, not people. Yours has 561 votes and 667,000 shares. So the ranking was never the point. Tonight is.
Tonight at 4:30 Central, Tesla either answers it in front of everyone, or skips it in front of everyone.
Model S and X owners, what year is your car, and what is it still waiting on?
$3.5 trillion on the table and Elon Musk cannot sit in the talks.
Tesla and SpaceX just moved their merger from a podcast tease five weeks ago to special committees of independent directors tonight. Musk said it himself on the SpaceX earnings call. Shareholders asked Tesla and SpaceX to explore it.
SpaceX is worth about $2 trillion. Tesla is worth about $1.45 trillion. Put Tesla and SpaceX together and it would be the biggest merger ever.
Musk runs both companies. That is exactly why he is recused from the talks as a conflicted party.
No terms agreed. No merger guaranteed. It could still die in committee. But if it survives, Tesla shareholders and SpaceX shareholders get the vote.
And Tesla reports earnings tomorrow afternoon.
692 questions went in for Wednesday's earnings call. The HW3 question is sitting at number 7.
Look at the six above it. Cybercab scaling. Unsupervised FSD dates. Buying a Cybercab. Robotaxi cost per mile. What an Optimus costs. When Optimus makes money. Every one of them is about the future Tesla wants to sell.
Number 7 is the only one from people who already paid. Tesla has said HW3 cars cannot do unsupervised FSD. The question asks when owners find out if the upgrade is free, and when Tesla shows a real plan. 363,000 shares are standing behind that question.
And look how close it is. The Optimus revenue question above it has 366,000 shares behind it. That is not a gap, that is a nudge.
Voting closes tomorrow at 4 PM Eastern. HW3 owners who hold shares, this is your question, and it is one spot from passing a robot. Does it climb tonight?
app.saytechnologies.com/tesl…
The full top 7 list, via @SawyerMerritt
xitter.cf/SawyerMerritt/status/2…
Here are the top 7 voted for questions by retail investors for @Tesla's Q3 earnings call this Wednesday:
1) How does the Cybercab fleet scale past Austin, and what’s the regulatory path outside Texas given NHTSA’s open safety review?
2) When can we anticipate FSD unsupervised in customer cars?
3) When will the CyberCab be available for customer purchase/ownership?
4) What’s the unit economics per paid robotaxi mile, and how does it compare to the thirty-cent target?
5) What’s the cost per Optimus unit, and what does the first paying customer look like?
6) When do you anticipate Optimus will begin generating revenue?
7) Tesla has acknowledged that HW3 vehicles cannot support unsupervised FSD. When will owners know whether they can upgrade their hardware for free, and when will Tesla provide a clear plan?















