problems and glitches

[HC_79]

Software is hard. Lucid appears to be a company mostly run by electrical and automotive engineers. Like most who don’t live and breathe software, they clearly underestimated the challenge. Just wasn’t emphasized enough in the company culture. I think they’ll turn it around, though. When every review is “This car is amazing to drive, but the software…” that has to register eventually. No one at Lucid is happy with the current state of the software, I imagine.

But don’t expect it to turn around overnight. It might be good enough in a year, but it probably won’t be great for much longer than that.

They completely underestimated it. Having a team of only 20 software developers for an EV that’s is so integrated with the software is a huge management fail. They clearly didn’t think it was a priority and now there’s a lot of pressure on that team of 20 to get things fixed.

 

[MoniputerLM]

Well if you listen to this, it will upset most of you.

Lucid Motors’ Mike Bell on Software-Defined Innovation - Ep. 174

Stream Lucid Motors’ Mike Bell on Software-Defined Innovation - Ep. 174 by The AI Podcast on desktop and mobile. Play over 320 million tracks for free on SoundCloud.

m.soundcloud.com

I have been saying that Lucid would fix most of the software by the end of the year. Now I think it might be a year or more. Very sad if true.

 

[MoniputerLM]

FWIW it took Tesla ten years to add waypoint capability to their homegrown GPS navigation software.

I didn't buy a Tesla. I bought what I thought was a well thought out and well-engineered, including software, car that was developed by Lucid management after distilling what was good and bad in the luxury car market and built after observing the shortcomings of other manufactures. So far, the software and reliability are a big fail in my eyes but sounds like Mike Bell would disagree.

 

[joec]

I made it through the whole episode but kept wondering what company he was working for. Just a tad off from reality. And that dream drive pro, wow!

No surprise, he worked at Apple for 16 years. The Reality Distortion Field is alive and well.

Seriously, the parallels between this interview and pretty much any interview with any Apple exec are remarkable. He learned his PR skills well.

This guy will never talk about the current state of affairs. It’s always about where they are headed, not where they are. The people who work under him probably roll their eyes every time they hear him speak.

Now, this is a far cry from an interview with Elon Musk. I didn’t hear anyone promising autonomous driving by the end of the year. But it’s clear there’s some shared DNA in the fundamental belief that they have something to add that’s unique and game changing.

This interview is a great reminder of what Lucid is at its core: a Silicon Valley company. It should be required listening for anyone considering this car.

Just look at where Lucid are located. They have no interest in making “another car.” They want to upend the entire automotive industry. Otherwise, what’s the point?

Now, a lot of folks, including me, love to make fun of that sort of Silicon Valley hubris. But you know what? That level of audacity is what ultimately does lead to the state of the art moving forward. (Makes me miss my old stamping grounds a bit.)

You have to challenge assumptions to change the world. And you have to dream big to do something great. And when you dream big, you sometimes fall flat on your face. Such is life. You pick up the pisces and move on.

People who think things are just fine the way they are generally don’t put a dent in the universe.

There might be a greater chance Lucid falls on its face than it succeeds. But the one thing Lucid won’t do is just plod along and deliver the status quo. If you prefer the status quo, there’s nothing wrong with that. But Lucid is probably not the car company for you.

Like I said. Required listening. If this makes you vomit, consider ordering something else. If you can look beyond the Valley speak and see underneath they are trying to improve the every day experience of driving, then you might just want to come along for the ride. Bumpy as it may be.

 

[dawktah LucidGT]

That's dated July 3rd but sounds like it was recorded in January or February.

 

[MPawelek]

They may be trying to “move the state of the art forward“ but in the mean time charging us regular folks a whole ton of our hard earned money for a inferior product. I wish corporate executives were paid by performance and results like they were years ago. In the past 25 years it appears many end up getting huge bonuses and golden parachutes for poor performance and sometimes outright failure.

 

[hmp10]

You have to challenge assumptions to change the world. And you have to dream big to do something great. And when you dream big, you sometimes fall flat on your face. Such is life. You pick up the pisces and move on.

I agree with this. The only problem as it applies to Lucid is what buyers were told over and over. Instead of saying something like, "we are going to produce a state-of-the-art electric vehicle that will be a superb driving machine with software that, over time, will continue to add to the ownership experience", we got, "we understand that at this price point and to attract customers from German luxury brands, we have to produce a product from the outset that delivers everything we have promised in terms of initial quality."

My brother has come to sit our house and dog while we are traveling. Today we set up a guest profile for him to drive the Lucid, and we found that he is unable to operate the garage doors unless we go through the hassle of programming them again for his profile. We have to sign in all over to Tidal for him to use it. Alexa will not respond when the car is on his profile.

It's ridiculous to require a half hour or more of programming of almost every damned software feature of the car just to set up a guest profile. He's decided he'll use the Tesla Plaid most of the time, as more of its features remained accessible under his guest profile.

I've been driving for over 50 years, and I've never had to deal with having to do so much work just to let someone else use the car's features when they drive it. If this is where dreaming of the future actually takes us, I'll take a pass.

 

[joec]

I agree with this. The only problem as it applies to Lucid is what buyers were told over and over. Instead of saying something like, "we are going to produce a state-of-the-art electric vehicle that will be a superb driving machine with software that, over time, will continue to add to the ownership experience", we got, "we understand that at this price point and to attract customers from German luxury brands, we have to produce a product from the outset that delivers everything we have promised in terms of initial quality."

My brother has come to sit our house and dog while we are traveling. Today we set up a guest profile for him to drive the Lucid, and we found that he is unable to operate the garage doors unless we go through the hassle of programming them again for his profile. We have to sign in all over to Tidal for him to use it. Alexa will not respond when the car is on his profile.

It's ridiculous to require a half hour or more of programming of almost every damned software feature of the car just to set up a guest profile. He's decided he'll use the Tesla Plaid most of the time, as more of its features remained accessible under his guest profile.

I've been driving for over 50 years, and I've never had to deal with having to do so much work just to let someone else use the car's features when they drive it. If this is where dreaming of the future actually takes us, I'll take a pass.

Absolutely. This is why CarPlay and Android Auto are so compelling to so many people. Sign in once with your phone, and all of your accounts and preferences are automatically available to any car.

 

[tkhan456]

I think this is where a lot of the disconnect lies with Lucid and expectations. Sure, Lucid could have built their entire platform with “off the shelf” components. They made a conscious decision not to. They (rightly, if you ask me) figured in a luxury space, especially in the long run, they needed their software to help with their brand differentiation. It’s a key company value. The entire driving experience has to be unique. Few of the physical components of the car are off the shelf, so none of the software would be, either. But that means way more time and effort needs to go into development. And that’s where something seems to have gone wrong.

Tesla made the same bet, and it worked out very well for them.

Once you build on a foundation of your own bespoke software, simply “plugging in” pre-built components and making them conform to your system often ends up taking as much time as just writing it yourself.

Short-term pain for long-term gain. Sucks for us early adopters in the meantime. But that’s why it’s good other car brands exist.

Eh. I like CarPlay a lot. I wouldn’t have considered lucid if they had no plans to bring it. I couldn’t care less for any automakers in house infotainment.

 

[n8236]

If he says Software Defined one more time, I’m gonna kill somebody. I work in a true SD network comms space and cars are not there yet

 

[BJR]

Software is hard. Lucid appears to be a company mostly run by electrical and automotive engineers. Like most who don’t live and breathe software, they clearly underestimated the challenge. Just wasn’t emphasized enough in the company culture. I think they’ll turn it around, though. When every review is “This car is amazing to drive, but the software…” that has to register eventually. No one at Lucid is happy with the current state of the software, I imagine.

But don’t expect it to turn around overnight. It might be good enough in a year, but it probably won’t be great for much longer than that.

amen-software is hard

 

[hydbob]

I agree with this. The only problem as it applies to Lucid is what buyers were told over and over. Instead of saying something like, "we are going to produce a state-of-the-art electric vehicle that will be a superb driving machine with software that, over time, will continue to add to the ownership experience", we got, "we understand that at this price point and to attract customers from German luxury brands, we have to produce a product from the outset that delivers everything we have promised in terms of initial quality."

My brother has come to sit our house and dog while we are traveling. Today we set up a guest profile for him to drive the Lucid, and we found that he is unable to operate the garage doors unless we go through the hassle of programming them again for his profile. We have to sign in all over to Tidal for him to use it. Alexa will not respond when the car is on his profile.

It's ridiculous to require a half hour or more of programming of almost every damned software feature of the car just to set up a guest profile. He's decided he'll use the Tesla Plaid most of the time, as more of its features remained accessible under his guest profile.

I've been driving for over 50 years, and I've never had to deal with having to do so much work just to let someone else use the car's features when they drive it. If this is where dreaming of the future actually takes us, I'll take a pass.

That's the limitation of AAOS. Which will not be fixed with CarPlay or Android Auto

 

[wdupre1]

As a retired software engineer I'm actually not surprised that the software can be twitchy. First off, Lucid didn't write all of it. They integrated vendor packages into their software framework. To put all the blame on Lucid is ridiculous. Debugging hardware rarely requires as much effort as software. Hardware failures are almost always physically observable and readily reproducible. That makes the diagnosis of the problem orders of magnitude easier. A software issue that occurs on a processor in a cluster of many interconnected CPU's, each running at tens of billions of operations per second, is a bit more difficult to catch.

Ask yourself these questions: How many years did it take Microsoft (or Apple) to make their operating systems reliable? Did the fact that early versions of MacOS or Windows would occasionally run off into the weeds stop you from using a personal computer?

I'm gonna keep driving my AGT. It's stunning to look at, comfortable, and fun to drive. I am optimistic that the software will improve through iterative refinement.

 

[dawktah LucidGT]

As a retired software engineer I'm actually not surprised that the software can be twitchy. First off, Lucid didn't write all of it. They integrated vendor packages into their software framework. To put all the blame on Lucid is ridiculous. Debugging hardware rarely requires as much effort as software. Hardware failures are almost always physically observable and readily reproducible. That makes the diagnosis of the problem orders of magnitude easier. A software issue that occurs on a processor in a cluster of many interconnected CPU's, each running at tens of billions of operations per second, is a bit more difficult to catch.

Ask yourself these questions: How many years did it take Microsoft (or Apple) to make their operating systems reliable? Did the fact that early versions of MacOS or Windows would occasionally run off into the weeds stop you from using a personal computer?

I'm gonna keep driving my AGT. It's stunning to look at, comfortable, and fun to drive. I am optimistic that the software will improve through iterative refinement.

Agreed, I've been using Windows since before 3.11. I can't imagine how many BSD I've had over the years and at the most inopportune times. Still using Windows on version 11.

Maybe it's me, I don't have locking issues. I refuse to use my phone as a key. To me that's giving too much control by a device that can be hacked.

Software needs work but in daily use is not that big of a deal. When I travel on the highway it's either to Chicago or Louisville where I already know my way around.

Issues I do have, I'm solving on my own ironically. Turn off Wi-Fi, car wakes immediately in app for example.

 

[MPawelek]

As a retired software engineer I'm actually not surprised that the software can be twitchy. First off, Lucid didn't write all of it. They integrated vendor packages into their software framework. To put all the blame on Lucid is ridiculous. Debugging hardware rarely requires as much effort as software. Hardware failures are almost always physically observable and readily reproducible. That makes the diagnosis of the problem orders of magnitude easier. A software issue that occurs on a processor in a cluster of many interconnected CPU's, each running at tens of billions of operations per second, is a bit more difficult to catch.

Ask yourself these questions: How many years did it take Microsoft (or Apple) to make their operating systems reliable? Did the fact that early versions of MacOS or Windows would occasionally run off into the weeds stop you from using a personal computer?

I'm gonna keep driving my AGT. It's stunning to look at, comfortable, and fun to drive. I am optimistic that the software will improve through iterative refinement.

I think you are missing the point. Yes software is complicated and hard to implement. No one is refuting that point. The point is we are being charged 139k-180k for a inferior product that actually works on vehicles that cost 1/3rd. the price. Work on software should have been started sooner, with a different team or the actual production of the cars stalled until they were at a semi competent level of operation. The Air does not just have a few minor glitches. A lot does not work properly.

 

[hmp10]

Ask yourself these questions: How many years did it take Microsoft (or Apple) to make their operating systems reliable? Did the fact that early versions of MacOS or Windows would occasionally run off into the weeds stop you from using a personal computer?

The difference with putting all this UI software in cars is that people are often stabbing at malfunctioning buttons and icons or trying to figure out workarounds while piloting through traffic or barreling down the highway in a two-and-a-half-ton vehicle.

I have seen this false equivalency between cars and smartphones or personal computers over and over. I'm sorry, but it's nonsense.

And don't even get me going about the accidents and near-accidents caused by people texting on their smartphones while pretending to themselves they're driving a car.

 

[mon40]

I've been reading the lists of Lucid problems posted on this forum and I'm astounded. I understand that a new company will undergo problems. All products have a few glitches that the manufacturer usually corrects in the first few months. (Except for Microsoft WIndows, of course.) I can't quite comprehend the quantity of faults posted by people driving $150,000 Lucid vehicles. How can Lucid send so many cars out the door with so many issues? I put down my deposit for an Air Pure last October. Lucid claims the Pure will start rolling at the end of this year. As much as I desire my Pure, I'd rather wait an extra few months for Lucid to get its act together. I don't want a car that I must reboot to change radio stations or open the garage door.

What's with the windows hate. Im actually buying my lucid in cash with msft stock that i bought at $7 in 20 years ago. Not to mention i have been programming with their tech since 1994 and have just retired at 55

 

[dawktah LucidGT]

Ok, what DOESN'T work? I hear complaints about the HomeLink. It works, the implementation is just cumbersome. C'mon its GIGO. Its a dumb computer outsmart it. Instead of trying to make profiles for each person put everybody on the SAME profile. So far I haven't seen a limit on how many facial recognition images you can take. I apologize if there's a limit.

I think you are missing the point. Yes software is complicated and hard to implement. No one is refuting that point. The point is we are being charged 139k-180k for a inferior product that actually works on vehicles that cost 1/3rd. the price. Work on software should have been started sooner, with a different team or the actual production of the cars stalled until they were at a semi competent level of operation. The Air does not just have a few minor glitches. A lot does not work properly.

I call the the "Tesla Effect" We now live in a society that we expect all cars to have a tablet glued to the dash and function like a Tesla. Sorry, I don't agree. Drive it like its a '57 Chevy or '65 mustang with just an AM radio. What did we pay for the cars handling, range, comfort or the circus?

The difference with putting all this UI software in cars is that people are often stabbing at malfunctioning buttons and icons or trying to figure out workarounds while piloting through traffic or barreling down the highway in a two-and-a-half-ton vehicle.

I have seen this false equivalency between cars and smartphones or personal computers over and over. I'm sorry, but it's nonsense.

And don't even get me going about the accidents and near-accidents caused by people texting on their smartphones while pretending to themselves they're driving a car.

Its a circus! We have become such a computer dependent society we cannot function without it. We were warned in 1984! "Push the button"

 

[MPawelek]

I have cars that are over 100 years old. They all start and run every time and get me where I need to go. I never get locked out, the dash boards do not disappear while driving, tires do not explode when I run over a pot hole, I do not get lost driving them and they all have spare tires. They are dependable all of the time and if repairs are necessary common hand tools are all I need. Your mileage may vary…..

 

[MPawelek]

PS- What is old is new again. The 1956 Messerschmitt had a steering yoke way before Tesla and the color for the 1925 Ford Coupe is 97 years before Zenith Red!