Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

A post nobody asked for: How I'd handle the FSD problem (features, HW upgrades) for early adopters

This site may earn commission on affiliate links.
Hardly anyone cares about 4 cameras versus 8.
The only people that even talked about differentiating features were obsessive Elon twitter followers, and Elon never said what those features were even going to be.
The AP2 driving coasts to coasts doesn't give people what they paid for. It would have been just another side show. I didn't pay good money to watch a damn demo.

I disagree.

A lot of us — I know I did — certainly bought FSD because we believed it would deliver us additional features over buying just EAP. The eight active cameras and later Elon’s tweets were an integral part of that calculus.

Were it possible to ask around back in 2016 and I guarantee you (almost) nobody would have thought three years later FSD got us nothing. Very few expected FULL self-driving to be here now. We expected return for our money in more conventional ways which was reasonable and promised by Tesla.

As for the coast-to-coast sideshow, missing that goes to additionally show how Tesla has failed to deliver the development speed they promised with a public timeline.

You can not say there was no timeline or no tangibles. Yes there were and Tesla has missed ALL of them regarding original FSD product.
 
Hardly anyone cares about 4 cameras versus 8.

I cared!! That is one of the main reasons I bought FSD! I wanted access to all my cameras. It was a hard differentiator between the packages at the time.

And for people who say I was an idiot for believing they could get EAP working with only 4 cameras, look how well AP 1 works with just 1 camera. It was easily believable that with just adding the side cameras EAP auto lane change would work. At the time I bought AP2 was still only using 1 camera, so no one really knew how that was going to play out.
 
Over the course of time it seems like most of us have accepted that the only true deliverable of FSD is HW3.

Agreed, I have come to this conclusion. For many of us early adopters, hardware 3 wasn’t a calculation at all because our cars supposedly had all the hardware needed. I remember when HW3 became a thing shortly after I bought and thinking, well, that sucks. I was happy when they said we would get the upgrade, but I also knew that meant I was NOT getting any FSD feature until that happened and folks with brand new cars who hadn’t been waiting would get the features first.


If I had bought my car after the HW3 announcement, I am not really sure what I would have done. I very well may have not bought FSD and just waited until upgrades were happening. I might have picked up the package in the fire sale for $2k.



I know that if I bought a new Tesla today, I would not be purchasing the FSD package. I use none of the features except auto lane change and that ain’t worth $7k. Since we now know the FSD package is heavily focused on city driving vs no-nag highway level 3 (what I thought they would get working first), I am not really that interested anymore.



To the original subject of the thread, I would be OK with the option for a refund of my package at the bare minimum. I would not want to be forced to take a refund, since that would mean Tesla could wash its hands of our cars. With the option I could choose when I want out. I honestly don’t think they will even do anything that simple. I am guessing because then they would have to admit they have a problem.

I am not as much of a fan of an FSD credit. Maybe I don’t want to buy another Tesla after all this mess.
 
Last edited:
I disagree.

A lot of us — I know I did — certainly bought FSD because we believed it would deliver us additional features over buying just EAP. The eight active cameras and later Elon’s tweets were an integral part of that calculus.

Were it possible to ask around back in 2016 and I guarantee you (almost) nobody would have thought three years later FSD got us nothing. Very few expected FULL self-driving to be here now. We expected return for our money in more conventional ways which was reasonable and promised by Tesla.

As for the coast-to-coast sideshow, missing that goes to additionally show how Tesla has failed to deliver the development speed they promised with a public timeline.

You can not say there was no timeline or no tangibles. Yes there were and Tesla has missed ALL of them regarding original FSD product.

We all had different reasons for buying it, but at the end of that day FSD was the core of the promise. Having the car drive itself was the overriding expectation, and what was promised. Having an additional feature that you wouldn't have had might make you feel better, but that wasn't the promise. Demonstrating that they were closer wasn't the promise either.

The 4 versus 8 camera seems important to you because to you because it's appears to be a concrete promise they made that they failed to deliver on from a very "is it active" interpretation. But, there is huge difference between fully utilizing the camera feed versus only using a quarter of the pixels. With HW2/HW2.5 the system simply can't use the full resolution of all the cameras. Plus what matters more than using all the cameras is being able to detect more objects with the cameras, and to detect them accurately.

I think most people are going to be 100% forgiving over technical details changing as long as a feature does what it was advertised as doing. Now obviously that's not true in this case so the next best is to simply be better in the spirit of what the claim was. I vastly prefer having more brains behind 8 cameras than simply having 8 cameras versus 4.

So from a "am I getting something more" standpoint I think they accomplished that for HW2.5 owners with FSD. Where they're anticipating the HW upgrade that should come within the next 6 months or so.

HW3 is ultimately about buying time. Time you have with your car where you'll continue to get SW upgrades as it slowly marches to the unattainable goal of FSD.

FSD is the promise, and that's all that matters at the end of the day.

it's both the magic of it, and the curse of it.

The magic of it because it's what makes it so appealing. The dream of the car taking over responsibility for the driving even if only on approved freeways.

The curse of it because that responsibility requires sign offs from people outside of Tesla who might never sign off on a HW3 level Tesla. That means there can't be a promised time as its outside of their control.

Tesla promised a destination, and instead delivered a journey. The big question is whether people will be happy enough with the journey to forget about the destination. With HW2 owners that time is almost up, and why it's so important to take care of them because of the lack of any progress during their ownership period.

I think the vast majority of owners regardless of their stated reasons for buying it won't truly be satisfied until they can put it into NoA, and read a book while there car goes down the freeway. That to me is the minimal sign off point.
 
I know that if I bought a new Tesla today, I would not be purchasing the FSD package.

This would be an easy decision if Tesla included HW3 in every Tesla shipping out the door, but as I understand it they haven't completely switched over to HW3. We have lots of reports from owners with cars delivered last month with HW2.5.

The problem with HW2.5 is it's basically failed every AEB, AP, TACC, etc test thrown at it.

I really enjoy my Mode 3, but I have to admit HW2.5 has been a complete and total let down. It certainly does more than AP1, but it's just not good enough even without considering FSD stuff.

WK057 probably summarized it best when he said it was simply too inconsistent. He was talking about AP/NoA, but it doesn't matter if its TACC, AP, NoA, Smart Summon. It's all an inconsistent mess.

IIHS pedestrian detection it didn't ace because it didn't detect the kid soon enough to stop all the way
The AAA pedestrian detection it completely failed because it was a more challenging test.

Having HW2.5 means you're at the mercy of Tesla, and whether they improve anything
Having HW3 means your very likely to get the upgrades that will fix the problems.

Having HW3 is vastly more important than FSD, but Tesla still ties the two things together at least on the Model 3.

I simply can't recommend a Tesla with HW2.5

There is no promise with Tesla with HW3, but at least they got time.

Its hard because the Model 3 is still the best EV for the money by far.
 
  • Informative
  • Like
Reactions: TM34M and Dr. J
We all had different reasons for buying it, but at the end of that day FSD was the core of the promise. Having the car drive itself was the overriding expectation, and what was promised. Having an additional feature that you wouldn't have had might make you feel better, but that wasn't the promise.

I disagree. Looking back very few people actually expected Tesla to deliver ”FSD” anytime soon in 2016. That was hardly a calculation as it seemed to be at least two+ years away anyway. What people were buying were eight active cameras. Go back and read Reddit, TMC, what have you and you will see people discussing this in 2016. Why pay 3k for something two years away? Because the 8 active cameras requirement was expected to bring features sooner. (And indeed it did, Tesla just gave them freely to all and sundry.)

We had seen AP1 come about feature by feature and we expected FSD to happen the same way — Tesla said when over 4 cameras became active, we would need FSD to enjoy that, so we bought it to get those features. Elon confirmed this interpretation in January 2017 when he was asked when these ”differentiating” features would come and he said 3-6 months.

In fact I fully expect this was Tesla’s plan too. What we did not know was how little Tesla had in 2016 and how much they played us for fools, so when Tesla only later made some progress, found out their plans would fail and they would have to dramatically recalibrate, it all changed. This we failed to anticipate.
Demonstrating that they were closer wasn't the promise either.

Yes it was. Tesla made many promises and you are simply trying to ignore some of them for some reason.
 
Folks, you have to remember the historical context of 2016.

Back in 2014 Tesla presented AP1 as ramp to exit highway system with meet you at the curb Summon. OK. That was the lofty promise. But it did not come overnight. What they lauched it with was: lane departure warning, NOTHING more. Then it came out feature by feature: TACC, Auto Steering, lane changes and so forth. These came out over months and months, one by one.

This is exactly what people then expected for ”FSD” in 2016. EAP was to simply replicate AP1, FSD would mean the next steps towards the next lofty goal, step by step just like AP1 had happened. We expected EAP would come out as one in December 2016 because that is what Design Studio stated and it was not much beyond AP1 either. Then we expected FSD pack to be released feature by feature over the next months and years.

That’s why many of us bought it. Not because we thought the full package would be out overnight ever. But to buy access to the journey of those eight cameras becoming active.

None of us other than perhaps the most negative of trolls expected Tesla would not deliver any similar steps towards the FSD pack over the next three years and that just buying EAP would have sufficed.
 
Yes it was. Tesla made many promises and you are simply trying to ignore some of them for some reason

Trivial promises that don't mean anything in the greater scheme of things.

The 4 versus 8 cameras is the only ammunition that you have, and I explained why that doesn't make any difference. Someone with FSD is still getting more sensing data (with HW3), and that accomplishes the spirit of it.

Back then FSD wasn't organized as a set of deliverables like it is today.

All the step by step deliverables were under the EAP column along with a promised date. A date they missed by years as it was supposed to be feature complete years ago. They just got done recognizing EAP money from years ago for accounting purposes.

For FSD there was an expectation that some things would come first like L3 freeway driving because that goes under the FSD column, and is the most logical extension from EAP. Its the very first true deliverable that someone is going to expect for FSD. There wasn't an expected delivery date for this.

The city NoA was expected to come much later. There is far more work to be done for City Autopilot than freeway so no one had an expectation for city based features to come first.

With EAP the expectation was NoA, but vastly better than it is today
With FSD the expectation was NoA, but without having to pay attention.

If you really think people would have been satisfied without at least L3 on the freeway then I question whether you really understand people. There is a big difference between the rational excuses we use to buy something versus the emotional reason we really bought them. The people would not have been satisfied with a throw a bone type feature. At most that would have bought them time, but it's not like they didn't have time since what else were we going to buy?

It's ABSOLUTELY ESSENTIAL to remain steadfast on L3 or bust. That this is truly what it's about.

Why?

Cause City NoA might be another lie to buy time.

You have to ask yourself why are we getting things that rely on stop light detection, stop sign detection, etc when we already have NoA on the freeway. Is it because NoA on the freeway is never going to get approval? Which also means City NoA also isn't going to get approval? Why hasn't there been any work to really nail down NoA on the freeway. Why is Smart Summon so stupid even with good map information?

Anything short of L3 puts the responsibility on the driver.

Putting responsibility on the driver is buying time, and falls far short of FSD.

L3 is still short of FSD, but at least it's no longer L2.

City NoA is so weird that autonomous driving experts can't wrap their heads around how it's going to work.
 
  • Love
Reactions: run-the-joules
I feel for the people that leased there car, but why add anything on a car you don't own that you can't take with you, they may never see FSD, I did ask the service center here in Kansas City, and they said they can upgrade HW2.0 to HW3.0, they have always did right by me.
 
@S4WRXTTCS

Since you keep ignoring the totality of my arguments on which I base my views, I will do the same. You are wrong though — especially so from the perspective of what was promised and what many bought in 2016. But you are also wrong about what Tesla is doing today.

People were absolutely expecting FSD to appear peacemeal and in the ways and timeline Tesla outlined in late 2016, early 2017. That is how AP1 appeared, peacemeal, and it was logical AP2 would appear the same way as well. Elon Musk confirmed this thinking by saying FSD differentiating features would come by summer 2017.

That Tesla failed to deliver this because they had to rework their entire NN and the AP2 team is simply telling of the fact they mislead us about their status in 2016 and failed to ”fake it until they make it” this time. The four vs. eight active camera story was lost in that debacle and AP2 owners became victims of Tesla’s unachievable promises.

And Tesla never discussed Level 3 that you brought from somewhere. They discussed Level 5. And as nobody expected Level 5 to come in a year or even two, many (we) expected peacemeal progress towards it. I would even wager most expected it — and indeed that is how FSD finally now is playing out, peacemeal... just like AP1 did.

And yes, Tesla provided timelines and tangibles they failed to deliver. You are just making stuff up.
 
I feel for the people that leased there car, but why add anything on a car you don't own that you can't take with you, they may never see FSD, I did ask the service center here in Kansas City, and they said they can upgrade HW2.0 to HW3.0, they have always did right by me.

For the same reason people lease cars: For the use value. Tesla said eight active cameras would require FSD pack on the Design Studio and that EAP would only use four.

Back in 2016 everyone expected peacemeal progress throughout the lease period with Elon Musk saying 3-6 months to the first features for the FSD option. People expected to be using features brought by the FSD pack by summer 2017.

That is how AP1 played out and it was a reasonable expectation back in 2016 of how AP2 would play out too, until we learned what fools Tesla took us for of course.

Many things look different today...
 
@S4WRXTTCS

Since you keep ignoring the totality of my arguments on which I base my views, I will do the same. You are wrong though — especially so from the perspective of what was promised and what many bought in 2016. But you are also wrong about what Tesla is doing today.

People were absolutely expecting FSD to appear peacemeal and in the ways and timeline Tesla outlined in late 2016, early 2017. That is how AP1 appeared, peacemeal, and it was logical AP2 would appear the same way as well. Elon Musk confirmed this thinking by saying FSD differentiating features would come by summer 2017.

That Tesla failed to deliver this because they had to rework their entire NN and the AP2 team is simply telling of the fact they mislead us about their status in 2016 and failed to ”fake it until they make it” this time. The four vs. eight active camera story was lost in that debacle and AP2 owners became victims of Tesla’s unachievable promises.

And Tesla never discussed Level 3 that you brought from somewhere. They discussed Level 5. And as nobody expected Level 5 to come in a year or even two, many (we) expected peacemeal progress towards it. I would even wager most expected it — and indeed that is how FSD finally now is playing out, peacemeal... just like AP1 did.

And yes, Tesla provided timelines and tangibles they failed to deliver. You are just making stuff up.

I was very active on here during the whole introduction of AP2.

There was no piecemeal aspect of FSD like AP/EAP. In fact people assumed it was a completely separate code base, and that accounted for the demo being so good.

If FSD was really going to come piece meal then Elon would have spelled out HOW the features would differentiate just like they do so today. We had great debates on TMC about what the features Elon was even eluding to when tweeted a response to when features would differentiate.

You are correct that Tesla never discussed freeway Level 3. But, that is the most logical stepping stop on the way to FSD. That is the piecemeal progress towards it.

AP1 was piecemeal in a way that made sense.

FSD is piece meal in a way that makes a person seriously scratch their heads wondering what in the world is going on.

To me this thread is about what would satisfy most people. I strongly believe most HW2.5 owners who have FSD would be satisfied with a more targeted approach to L3 on the freeway. Where Tesla made improvements to the point where it truly felt like it would pass regulatory approval, and that Tesla made everyone aware of the steps being taken to accomplish that.

I kind of think of it like a Steakhouse

We bought Steak, and we expect Steak.

You can't piecemeal steak. It's gotta be actual steak.

It can't be bread
It can't be garnish on the side
It can't be dessert

It can be a small piece of steak. heck it can be a teaser piece of steak to hold us over.

But, see the thing is that Tesla doesn't have any steak. They're doing this crazy piece meal thing where they give us everything except steak. They have one more feature for EAP people (self park) before they start really giving FSD owners stuff. But, not stuff from the old playbook, but the new playbook.
 
Last edited:
Why pay 3k for something two years away? Because of 8 active cameras.
How would it make any sense for Tesla to have two code branches, one for four cameras and one for eight cameras? There must have been people pointing out how ridiculous that was from the beginning.
I think the reason that Tesla hasn't released any FSD features is that they haven't figured out how to do so safely. Stop light and stop sign recognition is extremely dangerous if drivers are not vigilant. Can anyone propose another theory as to why they haven't released it? They have the code for it.
Nothing will satisfy buyers of FSD because they were promised something that it is simply not possible for Tesla to deliver. If FSD was level 3 on the highway I would definitely have bought it but I bet even that will be impossible for Tesla to deliver.
I hope people who leased FSD get a refund.
 
How would it make any sense for Tesla to have two code branches, one for four cameras and one for eight cameras? There must have been people pointing out how ridiculous that was from the beginning.

There will be two code branches for EAP and FSD, why couldn’t the cameras have been part of that? If they could have gotten NoA working with 4 cameras, I think they would have. EAP was never going to be hands off, eyes off, so the driver was always going to have to verify the lane changes in some manner.
 
How would it make any sense for Tesla to have two code branches, one for four cameras and one for eight cameras?
Nothing a simple config variable can't solve.

I do think the best way for Tesla to handle this going forward is to split FSD back into EAP & FSD. But in recent ER Musk refused the suggestion - because I think they are close to releasing City NOA (when they think they would have delivered all the functionality of FSD).
 
Nothing a simple config variable can't solve.

I do think the best way for Tesla to handle this going forward is to split FSD back into EAP & FSD. But in recent ER Musk refused the suggestion - because I think they are close to releasing City NOA (when they think they would have delivered all the functionality of FSD).

I don't think this makes any sense as EAP is problematic because it includes so many things that should be in the FSD column.

Like NoA on the highway was under EAP, but really belongs under FSD.

When they put EAP, and FSD into the same package they really cleaned things up nicely. It allows for one code base for all the difficult autonomous driving task (smart summon, smart park, NoA, NoA city, etc).

You heavily discount FSD for EAP owners (like they did) to get them over.

The new direction makes so much more sense than the old one. It's like they actually did the work this time to implement it. :p
 
There will be two code branches for EAP and FSD, why couldn’t the cameras have been part of that? If they could have gotten NoA working with 4 cameras, I think they would have. EAP was never going to be hands off, eyes off, so the driver was always going to have to verify the lane changes in some manner.

There are a lot of things that rely on more than 4 cameras because of how they are coded

There is safety elements even while manually driving, and one thing that separates out Tesla from other companies is their commitment to active safety even for people who don't buy the fancy SW
There is NoA for safer lane changes even though they still require driver verification
There is Smart Summon that has to watch out for pedestrians.
There is sentry mode, and dashcam which is proving to be a complete game changer so I expect additional camera footage to be available.

Why should an engineer be limited to description page written about a product BEFORE it was ever implemented? It was written before they designed it.
 
Last edited:
Why should an engineer be limited to description page written about a product BEFORE it was ever implemented. It was written before they designed it.[/QUOTE]

.

And that is the problem in a nutshell. They were selling us nearly finished code based on a video and also told everyone EAP would be delivered December 2016! It was all smoke and mirrors in hindsight, but Tesla sure put on a good song and dance. NOW we know they are capable of significant amounts of deception, but back then we were all a bit naive. Figured they were better than that. Turns out they weren’t. I might as well by a car from GM or Audi, Tesla is quickly getting to that trust level, and at least GM and Audi have good service
 
And that is the problem in a nutshell. They were selling us nearly finished code based on a video and also told everyone EAP would be delivered December 2016! It was all smoke and mirrors in hindsight, but Tesla sure put on a good song and dance. NOW we know they are capable of significant amounts of deception, but back then we were all a bit naive. Figured they were better than that. Turns out they weren’t. I might as well by a car from GM or Audi, Tesla is quickly getting to that trust level, and at least GM and Audi have good service

The problem is with Elon himself.

He's completely unwilling to compromise or to create a healthy working relationships with vendors. They used the Nvidia code base for FSD video. They could have built upon that codebase, but they were dead set on not being vendor locked. So they developed the entire ADAS code base themselves.

It was a miraculous job of the engineers to pull it off, but also a complete con that Elon played on the customers terms of time frame.

Ideally Tesla would get rid of Elon from all the crap he's pulled, but they can't.

So the next best is for competition to do what competition has a tendency to do. That is to make everyone a bit more honest.

The Electrify America charing network is pretty well built out so that should help a lot of people have more choices when it comes to an EV.

VW/Audi isn't a bad choice since the primary people responsible were removed, and arrested.
 
Customers with leases:
  1. Anyone who leased a Tesla with FSD and took delivery before 11:59pm on 12/31/2017 Pacific time gets free FSD for life (tied to the registered customer, not their account, with verification whenever a new vehicle is acquired to avoid shenanigans with accounts being transferred to others) on one leased Tesla at a time for as long as they maintain continuous leases (with reasonable exceptions for delivery delays, etc between cars). If they decide to buy one and stop leasing, it will also get FSD for free but that ends the gravy train.
  2. Anyone who leased a Tesla with FSD at delivery after people in (1.) but before the FSD price drop in February 2019 gets FSD for free on their next leased Tesla, and 2,000 miles of non-expiring Supercharger credit on both their current car and the next one.
  3. If a vehicle was delivered without FSD but the customer added it after delivery, the benefits applied are based on the time FSD was added, not when the vehicle was delivered.

  1. Yes. This is a fantastic idea, and it costs Tesla basically nothing.
  2. Another solid suggestion. And again, it basically costs Tesla nothing to make these people happy.
  3. Makes sense. Leased 2 years ago, bought FSD in January 2019, get the Jan 2019 award. Flawless.

Purchases:
  1. Anyone who purchased FSD for a non-leased vehicle in the same timeframe as Leases (1.) gets a lifetime FSD that can be used on one car at a time, with a process to transfer to a newly-purchased vehicle from Tesla (cannot transfer to used vehicles, nor between other vehicles they may already own).
  2. Anyone who purchased FSD for a non-leased vehicle in the same timeframe as Leases (2.) gets a one-time free FSD transfer to a new Tesla vehicle purchased any time before 1/1/2025, and 2,000 miles of non-expiring supercharging credit for both their current vehicle and the vehicle to which they add FSD.

  1. Again, this totally makes sense. They're an early adopter that ponied up before anybody else was even considering it, they deserve a token of appreciation.
  2. Once more, a solid suggestion. It doesn't seem like much until that person buys another car, and figures "Might as well buy another Tesla and get my FSD transfer". Throwing them the extra $7000 of value for free might even convince them to buy another before they migith have otherwise.

MCU1 customers:
  • Rumor on the street is that HW 2.5 + MCU1 vehicles may not be able to be upgraded to HW3
  • If an MCU2 upgrade is needed to implement HW3 and it can be done in a practical sense, they should do it. It's that simple, because it's the right thing to do. If they need to raise the FSD price for MCU1 cars that don't yet have it purchased, I think it's perfectly OK for them to do so, but those who already paid do so in good faith.

The problem I see with this is that while Elon says they're working on a MCU upgrade plan, it seems unlikely IMO. I don't think it's a direct replacement, but I could be totally wrong and thinking about the nav computer. But I was pretty sure he said it's not a drop in replacement. That could potentially be a significant complication. Otherwise, yes, if someone paid for FSD and the hardware in their car can't do FSD, then you owe them hardware that can do FSD.

HW2.0 customers:

We don't yet know what's going to happen with HW2.0 customers since they've got more than just computer differences to account for.

If it's determined that it's genuinely impractical to actually do the upgrade, which might require replacement of cameras, wiring harness, and god knows what else, Tesla should offer the owner their choice of one of the following:
  1. Full refund for FSD + interest + free FSD on their next Tesla
  2. An *extremely* generous trade-in credit towards a preowned HW2.5 or HW3 Tesla, which will include FSD and have the computer upgraded before delivery if needed. When I say "extremely generous", I mean it. They shouldn't be paying out of pocket for this, and the car they get should be quite a bit newer and better than theirs. If the customer does this, it'll be treated as if it were the car s/he took delivery of originally for date-based stuff above.
  3. A well-above-market-value-but-not-quite-as-totally-bonkers-as-option-2 trade-in credit towards a new Tesla, which will include FSD for free and any other benefits based on the date of purchase above.

Did you mean HW1? I thought the difference between HW2.0 and HW2.5 was somewhat minimal, and the HW2 and HW2.5 computers were connector compatible. Am I totally off on this? Anyway...

  1. Full refund, totally agreed. Interest? Not going to happen. Free FSD on next Tesla, absolutely yes. Even more so than any of the other groups above.
  2. If it's the case that HW2 can't be upgraded to HW3+FSD, and the HW2 car was a recent purchase (2018/2019), then I don't know if there's anything Tesla could do to make them whole except #1. If we calculate the compound interest on $6,000 over 3 years at 10%, it's only $8,000. IMO, that's basically the value of (1.) from each other the above lease/own sections.
  3. What about offering two free upgrade options, plus the free FSD, plus $5,000 off the price? Since we're again only talking around $8,000 lost value in a good case scenario, I feel like offering more than $8,000 compensation is pretty okay. Unless some amount of this is about punishing them for leading customers on for so long?