The Long Goodbye

[Need4Sleep]

It would be a miracle if Lucid can launch Cosmo in early 2027! Not talking about a paper launch. Consequential shipped to public volume. If they can ship 1,000+ REVENUE units by mid 2027, Lucid will be doing wel. But that still puts Lucid 1 year behind Rivian and 9 months or so behind BMW/Volvo.

Personally I wouldn't stress too much about being "first". Kia/Hyundai were how many years "late" to the EV game and yet they're making a killing now. What matters is that Lucid executes midsize well from the get-go.

 

[Adnillien]

It would be a miracle if Lucid can launch Cosmo in early 2027! Not talking about a paper launch. Consequential shipped to public volume. If they can ship 1,000+ REVENUE units by mid 2027, Lucid will be doing wel. But that still puts Lucid 1 year behind Rivian and 9 months or so behind BMW/Volvo.

Lucid would do well to follow their own timelines and commitments than to worry when the competition is releasing a car. Lucid may be behind schedule but it is not because Rivian or BMW released a car. The argument that Lucid is late because of Rivian R2 is like saying the Air was ten years late because of the Model S. It does not make sense.

 

[Leftlane]

If I already owned an Air or Gravity I wouldn’t want to see the Cosmos launch any time soon. It’s already hard enough for get a service appointment, isn’t it?

 

[BS8899]

Lucid would do well to follow their own timelines and commitments than to worry when the competition is releasing a car. Lucid may be behind schedule but it is not because Rivian or BMW released a car. The argument that Lucid is late because of Rivian R2 is like saying the Air was ten years late because of the Model S. It does not make sense.

I agree with your notion that Lucid should take its time to properly develop a quality mid-size vehicle before releasing it.

That said, competition is also real. Lucid will be late.

But, releasing a model that's not read y for prime time just to try to match the competition's timeline is ALWAYS THE WRONG ANSWER!

Lucid must stay the course. Do it right. No paper launches. None of those circus tricks will advance Lucid's cause.

 

[borski]

I think @Rogue has a point.

I am no AI expert, so I won't opine on the job description.

Here are my concerns:

> every platform at Lucid is running out of compute horse-power and in need of HW upgrade
> Cosmos debut is supposed to be Q4'26

I think many of us agree the SW is the weaklink in Lucid's product lines. BUT in the time frame of interest, is Lucid taking the right path to correct these problems?

AI is immensely powerful, and if harnessed correctly, can help Lucid's software organization and engineering immensely.

The key descriptor being harnessing it correctly.

 

[borski]

You are probably just a very, very patient person. What does your wife say about that?

When I buy stuff, no matter the price, I just want it to work how it's supposed to work. If I have to go to a forum to figure out some things, it's not working out for me lol. Again, life is too short to figure out the small nuisance of the car.

I have ADHD. I am extremely impatient.

 

[Nunya]

Personally I wouldn't stress too much about being "first". Kia/Hyundai were how many years "late" to the EV game and yet they're making a killing now. What matters is that Lucid executes midsize well from the get-go.

For mid-mid-size, it will follow the Gravity model - sell 2 or 3 on Dec 30, 2026 to say "we met the 2026 goal". First real deliveries in May '27, my guess.

BUT, they also said mid-size will be with the stated goal of being SSAE L4 capable. Make it L4 functional on day 1, on interstates, daytime only with promises of more capability to soon come. It would sell more than they could make.

 

[hwy17er]

What part do you disagree with?

There are multiple independent threads with fob issues. And, when someone is struck, there does not appear to be a common cure. Some folks reported improvement with a new fob. Others did not. Others suspect software. Some suspect hardware inside the car. Even the Lucid guy acknowledge the fob problem. Probably the majority of people do not have the problem. But it is clearly a material risk. As I noted above, there is nothing comparable to this on any of the other forums I am following for next car research. I am very familiar with adverse selection bias. But when you see a forum with more problem posts than lifestyle posts, that highlights major problems. The Cadillac Lyriq forums still look like this - probably half the posting activity are issues. And the Polestar, that had severe GHCA failures that dominated the forums for over a year, have largely settled into lifestyle posts.

My personal priority is that core dependability is highest. The Gravity has a material risk (we don't know if it is 1%, 10%, 50%, etc.) that fobs will be unreliable. I am not willing to take a chance that I could draw a lottery ticket and get stuck with a car I cannot depend on.

To Clarify: "the fob issue is a material percentage of owners" is the part I disagree with. It may be a material percentage of those posting on here but, and it's just my opinion that may be totally wrong, but I'm having trouble believing it's a large % of all Gravity owners...many of whom are not even on this forum.

That's it. I don't mean to devalue anyone's experience with the fob at all...in fact it seems like more posters reported fewer problems when the last couple substantial updates came out but then there was another contingent that had an increase in fob issues after 3.5.1 was released as discussed eloquently by @hmp10 following his phone conversation with Emad and that group seems to all have highly spec'd configurations, thus the processor overwhelming scenario.

 

[BS8899]

AI is immensely powerful, and if harnessed correctly, can help Lucid's software organization and engineering immensely.

The key descriptor being harnessing it correctly.

Exactly, "Harnessing it correctly".
The concept of "AI" is more than 75 years old. Now, we are at the fringe of having the HW/SW power to instantiate "AI". in a productive way. That said, it will take more time and research before it is truly useful in things like cars, airplanes etc..
Meanwhile, my vacuum cleaner is "AI". My washing machine is "AI". my microwave oven is "AI". I feel better already!

 

[hmp10]

To Clarify: "the fob issue is a material percentage of owners" is the part I disagree with. It may be a material percentage of those posting on here but, and it's just my opinion that may be totally wrong, but I'm having trouble believing it's a large % of all Gravity owners...many of whom are not even on this forum.

That's it. I don't mean to devalue anyone's experience with the fob at all...in fact it seems like more posters reported fewer problems when the last couple substantial updates came out but then there was another contingent that had an increase in fob issues after 3.5.1 was released as discussed eloquently by @hmp10 following his phone conversation with Emad and that group seems to all have highly spec'd configurations, thus the processor overwhelming scenario.

I've stayed quiet about this thus far, because I don't want to get anyone in trouble. But I can say this much . . .

I've dealt with Lucid mobile techs in three states due to issues we've had with our Lucids on road trips. All were extremely helpful, and all have been in their jobs since early days with the Air and dealt with its significant teething pains.

I have been told the whole mobile tech squad had high hopes that the Gravity introduction would go more smoothly than the Air. Instead, they have been very troubled and somewhat exhausted by the high volume of Gravity service calls, often related to issues that trace back to software. They characterize the Gravity rollout as even more challenging for them than the Air rollout, in part because they had high expectations that it would go more smoothly.

It's one of the reasons I have trouble with the notion that the vast majority of Gravity owners have experienced few or no problems. That is not the experience of the mobile tech community. One told me that he had 21 Gravities in his service area, and he has had multiple calls to almost every one, often for help with issues originating from software.

This does not mean I doubt reports from some here that their cars are relatively problem free. (Our second Gravity was for the first few months until 3.5.1 came along.) What I doubt is the reliability of extrapolating their good experiences to the wider ownership community on the supposition that happy owners just don't come to this forum to post about their experience the way complainers do.

 

[scal_air]

A lot of the service overload is tied to poor software on the Gravity. Most people don't check forums when their car isn't working how it should, so they'll call service for software-related issues. I'd be willing to bet that wait times improve dramatically if they make meaningful improvements on software in the next couple of months.

Oh no doubt. But there are plenty of software related issues that can't be solved by a CS call. But in the end, if a car isn't working as it should, is it better that it's a software issue versus a hardware issue??!! If Lucid could improve software reliability, it would certainly help the service centers and customer service. But the reality is that cutting service personnel at this point is exacerbating the problem and losing customer loyalty.

 

[dbsb3233]

If I already owned an Air or Gravity I wouldn’t want to see the Cosmos launch any time soon. It’s already hard enough for get a service appointment, isn’t it?

While true, of bigger importance is the company staying in business for the following 3... 5... 10 years that we may own our cars. A mass-market volume seller is necessary for that. Probably as soon as they can get it.

As annoying as a long service wait is this year, far worse would be no service at all in 5 years because the company goes under.

 

[allenmhc]

I hate to be the bearer of bad news, but if using agentic AI slop is their answer to stronger quality controls, everyone is in for a rough ride.

This is only on one job posting. However, AI slop is the absolute worst case scenario for a car handling systems that must work every time without fail.

The number of "new" platforms discussed is concerning over the past few years. You need at least two given the Air is on a dual processor setup and the Gravity a single processor.

Given they are hitting compute limits on relatively new hardware... I can't even wrap my head around how you do that unless performance is the lowest priority. Before folks go pre-optimization or performance but not stability, you would be astounded by the simplicity of the solutions, stability of those solutions, and the testability that are natural byproducts of prioritizing performance.

My hope is that they would clamp down on the AI slop and legitimately double down on a platform to make it work. Constantly shifting the goalposts and introducing the slop is legitimately the quickest way to derail a codebase.

This is a bad take.

Since the start of the year, every software company and dev shop has been using AI to help with software development. The reason everybody has jumped onboard is that it *has* proven to increase the speed, whether through security/compliance testing, feature development, or automated testing (note that the job posting mentions these use cases). They're not looking to hire non-engineers vibe-coding interfaces.

I mentioned this in another thread a few months back, but when I think 3.4 rolled out, they actually rolled back or reduced some of the fancy design elements in UX 3.0: the drive modes no longer animate, the seats are static images, the side cameras run smoothly w/o some sort of crop or filter through them. I'd bet that the problem wasn't the quality of each element (since each screen, by itself, was pretty well-built), but that there wasn't enough integration testing, when all the software pieces came together and all of them are stacked on top of each other. Hence, the new architecture and more of this type of full, end-to-end integration testing, and I suspect the 30% perf. improvement is work precisely to make the full stack be more efficient.

 

[exxxviii]

To Clarify: "the fob issue is a material percentage of owners" is the part I disagree with. It may be a material percentage of those posting on here but, and it's just my opinion that may be totally wrong, but I'm having trouble believing it's a large % of all Gravity owners...many of whom are not even on this forum.

That's it. I don't mean to devalue anyone's experience with the fob at all...in fact it seems like more posters reported fewer problems when the last couple substantial updates came out but then there was another contingent that had an increase in fob issues after 3.5.1 was released as discussed eloquently by @hmp10 following his phone conversation with Emad and that group seems to all have highly spec'd configurations, thus the processor overwhelming scenario.

Well... The number of people posting here with Gravity fob issues is greater than all the other forums I follow combined (Lyriq, iX, Polestar 3, and Rivian). I do not know if it is 1%, 5%, 10%, or whatever. But I do know that many people are dealing with this. And it is a far greater risk than I am willing to take, when any other car would be 100% dependable.

I do not think it is greater than 50%. If it were that high, then Lucid would have likely encountered it in testing and would have been negligent in releasing the car. The bigger problem than just having multiple contributors with fob issues is that there does not appear to be a cure. When GM had 12V failures with the Lyriq, that probably did affect 50% of owners. It was a software problem that took GM about 6 months to release, and then the problems were immediately solved. When Polestar had GHCA failures, that too probably affected a large percentage of owners. But it was a hardware problem and the latest part cured it. It does not appear that Lucid has any clue what fixes the fob issues.

Edit: The other thing I would add about the fob problems is that some people have it chronically with their cars. That is, if they have the fob problem, then it fails to function often. Other people post one-off occurrences of a fob issue that goes away the next time they drive the car. That makes it seem like there are multiple causes to the problem, bigger than just "software."

 

[LVAGT]

To Clarify: "the fob issue is a material percentage of owners" is the part I disagree with. It may be a material percentage of those posting on here but, and it's just my opinion that may be totally wrong, but I'm having trouble believing it's a large % of all Gravity owners...many of whom are not even on this forum.

That's it. I don't mean to devalue anyone's experience with the fob at all...in fact it seems like more posters reported fewer problems when the last couple substantial updates came out but then there was another contingent that had an increase in fob issues after 3.5.1 was released as discussed eloquently by @hmp10 following his phone conversation with Emad and that group seems to all have highly spec'd configurations, thus the processor overwhelming scenario.

In the past two years I've owned the Air, I've met a total of 8 other Lucid owners in person, not from the forum, and they all have problems with either their Air or the gravity. In fact, I'd love to swap a vehicle for a week to see this "problem-free" Lucid and actually believe a person on the forum, or is it that their tolerance is way higher?

You have to understand, most people, myself included, just learn to accept the problems after a while without ever going back on the forum. You can't exactly return the vehicle you bought; most just try to justify what they have bought. I get on here occasionally whenever I'm bored, and because I've found @hmp10 to be a very respectable person on here, that's why I even posted. His opinions are usually not biased like some of the other members. For a significant portion of the population, they do not go to any forum for problem-solving, just straight up CS. In the past 2 years, the service times went from 3 to 5 days to 2 to 3 weeks, and now close to more than a month and a half. My last service was 6 months ago; it took them over 2 months to fix a simple issue, because they are so backed up with the Gravity.

I wrote this at 3:51 am PST, that's how serious it is

 

[Rogue]

This is a bad take.

Since the start of the year, every software company and dev shop has been using AI to help with software development. The reason everybody has jumped onboard is that it *has* proven to increase the speed, whether through security/compliance testing, feature development, or automated testing (note that the job posting mentions these use cases). They're not looking to hire non-engineers vibe-coding interfaces.

I mentioned this in another thread a few months back, but when I think 3.4 rolled out, they actually rolled back or reduced some of the fancy design elements in UX 3.0: the drive modes no longer animate, the seats are static images, the side cameras run smoothly w/o some sort of crop or filter through them. I'd bet that the problem wasn't the quality of each element (since each screen, by itself, was pretty well-built), but that there wasn't enough integration testing, when all the software pieces came together and all of them are stacked on top of each other. Hence, the new architecture and more of this type of full, end-to-end integration testing, and I suspect the 30% perf. improvement is work precisely to make the full stack be more efficient.

I have no issues with non-generative AI. However, the generative AI has significant issues.

I have personally watched it forced onto developers who then completely outsource their critical thinking. Even for things that are supposed to be "trivial" like building a website, a plethora of errors and inconsistencies crop up. Given their tech stack is C++ mostly, the coding assistants and agentic AI workflows are particularly ill equipped for the nuance and complexity of the language. The cognitive bias effect of AI's expertise in areas you are unfamiliar with is particularly strong and well documented.

Open source has drowned, and amongst the hype are numerous studies and research demonstrating a marked decline in quality:
2020-2024 analysis of code repos: https://www.gitclear.com/ai_assistant_code_quality_2025_research
MIT study on cognitive impacts of AI usage: https://economics.mit.edu/sites/def...Cognition and Knowledge Collapse 02-20-26.pdf
METR's first and second study comparison: https://metr.org/blog/2026-02-24-uplift-update/ (Note the selection bias heavily cited in the second study and code quality differences)

My personal experience has been witnessing the neutering of engineers of nearly all critical thinking skills and the ability to reason about new areas outside their expertise. Engineers previously capable of figuring out errors by themselves handicapped to the point of inepitude and requiring assistance to resolve issues when AI couldn't figure it out for them.

I would argue it is a reasonable take grounded in the realities of the technology's limits and effects on developers both from third party studies, observation of impacts at a Fortune 200, and personal use.

 

[hmp10]

In the past two years I've owned the Air, I've met a total of 8 other Lucid owners in person, not from the forum, and they all have problems with either their Air or the gravity.

This is, of course, anecdotal. But a few weeks after 3.5.1 came out, the guy who does the "UpcurrentEVs" YouTube channel called us to meet him for dinner. He lives near us and has a Gravity he has been reviewing on his channel. He was preparing a video on software updates and wanted to compare notes with us beforehand, especially as we have two Gravities.

As we were looking over the menu and before I said a word about the Gravity, he suddenly asked if our cars had gone a bit haywire after the 3.5.1 update. He said his had, and he recounted some of the new bugs he was seeing -- some of the same ones we were seeing.

At least it made me more comfortable that the issues weren't unique to our cars, something I have wondered from time to time.

 

[martinjlm]

...

What-if they ally with Rivian on SW for the next generation? I know that's heresy for the Lucid enthusiasts.....but!!

If GM and Honda can collaborate on engines, transmissions, fuel cells and EVs...if GM and Ford can collaborate on transmissions (several times)...If BMW and Toyota can collaborate on a sports car... I could go on but hopefully you see my point.

 

[TahoeMichael]

I am sure @hmp10 is getting the VIP treatment because of his stature (owns multiple Lucid cars, knowledgeable, founding member of this website etc.). I think that's all fine. Hopefully, Emad/Lucid's response benefit not just @hmp10 but awaken Lucid to embrace the broader set of issues with their products in the field.

Without exaggeration, Lucid's survival is on a knife's edge. The plethora of issues with their current products and services and late to the game with mid-size, etc.. Lucid must course-correct to survive!

The next 6-8 months will delineate Lucid's fate!

I don't know that Lucid needs a wakeup call. They've already swept away a significant fraction of their leadership, including the CEO. The question is just whether they'll successfully navigate through the reset they've already started.

 

[TahoeMichael]

I hate to be the bearer of bad news, but if using agentic AI slop is their answer to stronger quality controls, everyone is in for a rough ride.

This is only on one job posting. However, AI slop is the absolute worst case scenario for a car handling systems that must work every time without fail.

The number of "new" platforms discussed is concerning over the past few years. You need at least two given the Air is on a dual processor setup and the Gravity a single processor.

Given they are hitting compute limits on relatively new hardware... I can't even wrap my head around how you do that unless performance is the lowest priority. Before folks go pre-optimization or performance but not stability, you would be astounded by the simplicity of the solutions, stability of those solutions, and the testability that are natural byproducts of prioritizing performance.

My hope is that they would clamp down on the AI slop and legitimately double down on a platform to make it work. Constantly shifting the goalposts and introducing the slop is legitimately the quickest way to derail a codebase.

I actually think the job posting is pretty reasonable, given the current strengths and weakness of AI. Nothing in the job posting talks about writing code that's included in the final product. It's all about test automation, code analysis, test coverage, etc. That's a place where AI can increase the productivity of the team without it's failures directly negatively impacting the code quality of the shipping product.