An Air UX 3.x Question for you Software Pros...

[Buffalo Bob]

I understand why it makes total sense to get the Air on the new platform in order to to ease the burden of implementing at least some future improvements and performance tweaks across Lucid's entire product portfolio. My question is about how disruptive the initial platform change might be. As I understand it, even before AI-driven programming, a lot of the change from one platform to another could be handled pretty seamlessly by the computer systems themselves, with little human intervention. With AI, I imagine that shift to machine-based change already is and ever will be even more dramatic. So my question is... When Lucid implements UX 3.0 for the Air, and let's assume that there are no changes other than taking the current UX 2.x.x software, and moving it to the new platform, how likely is it that the platform change alone would introduce a whole lot of bugs that did not exist in the old platform? Could it possibly be wildly disruptive at the onset, or more likely needing just a handful of tweaks, such as what inevitably happens when any such complex system is modified? Curious about your take on this.

 

[doubletwist]

It certainly has the potential to be fairly disruptive.

------
NOTE: I have no special insider knowledge of Lucid's software. This is based on public information I've seen, plus 30+ years of experience in IT as a Sysadmin mostly. I'm not a 'real' developer, but I've programmed a bit and have beta tested a lot of software. But I could be totally wrong.
------

I think them calling it "UX" 3.0 is a little misleading. It's not just a matter of throwing a different user interface design over existing back-end functionality. By the same token, because of the compute hardware differences between the Air and the Gravity, it's not just a matter of taking the Gravity Code and tweaking a few back-end bits.

The way 3.x works underneath is totally different from 2.x. From what I've read, they've changed the way they virtualize different parts of the system, in addition to consolidating into a single compute system rather than having separate hardware for the drive systems and infotainment. So now for the Air, they have to ensure that the new virtualization/hypervisor setup works across those two disparate systems.

There's a LOT of areas in those hardware/software architecture differences th where significant bugs could be introduced.

Idelaly, when designing 3.x for the Gravity, they tried to architect it in a way that is as modular as possible, to minimize the changes needed to transition to different hardware underneath, but given the recent turmoil in the software teams, it doesn't suggest the best design went into the software.

Hopefully, they are smarter about it for 4.x moving forward.

 

[Bobby]

I’d like to see the Gravity dashboard on the 2027 Air. The shape makes the display way roomier.

 

[AirBo]

I’d like to see the Gravity dashboard on the 2027 Air. The shape makes the display way roomier.

I'd like to see Gravity Software improvements on any Lucid GT Air, I would also like to see things they have promissed (battery to be used for home electricity outage) and many other things. Let us keep fingers crossed. Software is terrible. "Dream Drive" remained just a dream and more and more

 

[Bobby]

I'd like to see Gravity Software improvements on any Lucid GT Air, I would also like to see things they have promissed (battery to be used for home electricity outage) and many other things. Let us keep fingers crossed. Software is terrible. "Dream Drive" remained just a dream and more and more

I wouldn't quite say it's terrible. I do think there are some major features that they should implement.

 

[bobr]

how likely is it that the platform change alone would introduce a whole lot of bugs that did not exist in the old platform?

De-bugging is the process of removing/correcting bugs in software. Programming is the process of generating bugs in software.

 

[joec]

I understand why it makes total sense to get the Air on the new platform in order to to ease the burden of implementing at least some future improvements and performance tweaks across Lucid's entire product portfolio. My question is about how disruptive the initial platform change might be. As I understand it, even before AI-driven programming, a lot of the change from one platform to another could be handled pretty seamlessly by the computer systems themselves, with little human intervention. With AI, I imagine that shift to machine-based change already is and ever will be even more dramatic. So my question is... When Lucid implements UX 3.0 for the Air, and let's assume that there are no changes other than taking the current UX 2.x.x software, and moving it to the new platform, how likely is it that the platform change alone would introduce a whole lot of bugs that did not exist in the old platform? Could it possibly be wildly disruptive at the onset, or more likely needing just a handful of tweaks, such as what inevitably happens when any such complex system is modified? Curious about your take on this.

This is a great question. And the answer no one will like?

It depends.

There's certainly a lot of risk implementing this. There's the obvious risk of bugs. Then there's the risk owners will be so unfamiliar with it that they just decide to hate it straight out of the box.

Gravity owners who also happen to have an Air will likely be the fastest to approve and benefit. The rest of us? Might get lots of new threads here complaining about "Safety Issues" and "Why did they have to move this?" etc.

And at least one thread begging Lucid to put 2.0 back on their car.

Did I mention bugs?

They call it "UX 3.0," but it's as much an under-the-hood change as the cosmetic layer up top, I would think.

The good news is Gravity itself is far more bug-free than it was just a few months ago. But that doesn't mean running the same "guts" of the Gravity stack on the Air's older chipset might not cause a whole new set of issues to hammer out. Then there's the obvious screen size differences. The need to split the top screen into three distinct panels. The completely bonkers way the computers inside Air talk to each other.

Not to mention, lack of programmable buttons. Completely different steering wheel input devices.

The only real question is whether Lucid is willing to delay this if need be in order to be sure as many of the bugs as possible are worked out BEFORE launching it. No matter how much people here complain it hasn't been released yet.

For the love of Pete, Lucid, don't give anyone a hard date for when this thing is going to ship!

Ship it. Then two weeks later, tell us you shipped it. If any one asks in the meantime, just say it's still a work in progress. Even if you've already shipped it.

 

[Rogue]

The odds of an initial launch not having a swath of bugs revealed and perhaps even halting a rollout is very low. Speaking from experience, you can test and test and test until you think you've tested everything that can be tested.

Then you launch or release a major version for customers to break within the day. So you add more tests. And repeat.

The speed of fixes will be a far better indicator than anything else of the underlying quality. Solid foundations means easy pinpointing and rapid resolution. A slapped together over engineered monstrosity less so.

 

[SaratogaLefty]

Most likely there will be a group of "beta testers" chosen and perhaps might even already be running the UX3.0?? In any case, hopefully by the time the masses get it there won't be any significant surprises.

 

[northernstar]

Haven't kept up with the latest Air news/updates for a long time... and finally saw a few clips of the 2026 Investor Day presentation...
Is there an another thread that talks about Air's 2026 and 2027 software roadmap?

Impressions that I'm getting from the presentation are

• Another Center Console Control (CCC) upgrade on MY '28?
  • Can we just upgrade the '28 spec CCC instead of MY '24 ~ '27 version then?
• UI/UX 3.X probably the last significant software update for MY '27 and earlier Airs
• Sad to see no hands-free City driving coming to Air

 

[knewman]

My thoughts on software, coming from ~20 years in IT:

We like to imagine software development as an ordered, elegant, hierarchical process that results in a predictable and understandable software system. And this is true for software projects that have had the brightest minds and/or nation-state budgets applied to them, like aerospace applications, or something John Carmack was tapped to build.

But it's not the case for most software you or I use. The software we typically use is a rickety house of playing cards, stuck together by glue, spit, or anything sticky that happened to be close by. Not just 52 cards, but hundreds of cards, mostly from different decks that have been poorly maintained. The underpinnings of the software cobbled together by people that will likely never actually use the software, don't really understand how the many of the individual playing cards work, and have little personal stock in the end result beyond a paycheck.

So my answer to your question regarding how disruptive 3.0 will be? Very. Prove me wrong Lucid!

 

[Lucken]

My thoughts on software, coming from ~20 years in IT:

We like to imagine software development as an ordered, elegant, hierarchical process that results in a predictable and understandable software system. And this is true for software projects that have had the brightest minds and/or nation-state budgets applied to them, like aerospace applications, or something John Carmack was tapped to build.

But it's not the case for most software you or I use. The software we typically use is a rickety house of playing cards, stuck together by glue, spit, or anything sticky that happened to be close by. Not just 52 cards, but hundreds of cards, mostly from different decks that have been poorly maintained. The underpinnings of the software cobbled together by people that will likely never actually use the software, don't really understand how the many of the individual playing cards work, and have little personal stock in the end result beyond a paycheck.

So my answer to your question regarding how disruptive 3.0 will be? Very. Prove me wrong Lucid!

Well there’s some optimism to start my day.

 

[Mr.Quindazzi]

My thoughts on software, coming from ~20 years in IT:

We like to imagine software development as an ordered, elegant, hierarchical process that results in a predictable and understandable software system. And this is true for software projects that have had the brightest minds and/or nation-state budgets applied to them, like aerospace applications, or something John Carmack was tapped to build.

But it's not the case for most software you or I use. The software we typically use is a rickety house of playing cards, stuck together by glue, spit, or anything sticky that happened to be close by. Not just 52 cards, but hundreds of cards, mostly from different decks that have been poorly maintained. The underpinnings of the software cobbled together by people that will likely never actually use the software, don't really understand how the many of the individual playing cards work, and have little personal stock in the end result beyond a paycheck.

So my answer to your question regarding how disruptive 3.0 will be? Very. Prove me wrong Lucid!

Thanks for the insight. I do think this does apply to cars like the Lucid and has become the norm in these new "software defined" vehicles.
Not sure it applies to older vehicles equipped with say Bosch systems. That stuff was/is pretty robust.
I am always wondering if the trade off is worth it

 

[Buffalo Bob]

I’d like to see the Gravity dashboard on the 2027 Air. The shape makes the display way roomier.

I'd like to see Gravity Software improvements on any Lucid GT Air, I would also like to see things they have promissed (battery to be used for home electricity outage) and many other things. Let us keep fingers crossed. Software is terrible. "Dream Drive" remained just a dream and more and more

Haven't kept up with the latest Air news/updates for a long time... and finally saw a few clips of the 2026 Investor Day presentation...
Is there an another thread that talks about Air's 2026 and 2027 software roadmap?

Impressions that I'm getting from the presentation are

• Another Center Console Control (CCC) upgrade on MY '28?
  • Can we just upgrade the '28 spec CCC instead of MY '24 ~ '27 version then?
• UI/UX 3.X probably the last significant software update for MY '27 and earlier Airs
• Sad to see no hands-free City driving coming to Air

These comments are all very worthy of discussion, but frankly not 'on topic' for this thread. As the OP, I would much appreciate it if you would consider continuing them either on a new thread or an appropriate existing thread. Thanks!

 

[borski]

Thanks for the insight. I do think this does apply to cars like the Lucid and has become the norm in these new "software defined" vehicles.
Not sure it applies to older vehicles equipped with say Bosch systems. That stuff was/is pretty robust.
I am always wondering if the trade off is worth it

Lucid uses plenty of “Bosch systems” (and others). For example, the original TCS on the Air was third-party.

It couldn’t keep up with the Sapphire’s needs (or the lower-trim Air when it hits a slick patch with regen on), until Lucid built their own 1000Hz TCS for the Sapphire, which has now trickled down to Gravity and new models.

Say what you will, but Lucid can build great software, often beating their competition.

It’s their infotainment stack (and maybe comm stack) that needs a lot of help, imho. Not “all software in the car.”

 

[knewman]

Thanks for the insight. I do think this does apply to cars like the Lucid and has become the norm in these new "software defined" vehicles.
Not sure it applies to older vehicles equipped with say Bosch systems. That stuff was/is pretty robust.
I am always wondering if the trade off is worth it

Yes, totally agree, I should have qualified my thoughts in that they only apply to modern complex software, not a mature product with an arguably more narrow scope like a Bosch controller.