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

I'm In Seattle And As a Die Hard Tesla Fan, I am Not Sold On FSD Beta Experience. (Been On Beta since it rolled out to Public). Thoughts?

This site may earn commission on affiliate links.
I have been seeing an increasing number of tweets by fans and FSD beta users, liked and retweeted by Elon himself, touting the latest release of FSD Beta and the improvement it has made. However, I can barely go a few miles on even suburban streets without overtaking entirely to prevent something bad from happening.

I live in Seattle area, where the roads are smaller and compact, often times not the big lanes with big sidewalks and paved roads that CA has.

Washington has potholes, small lanes, cars parked on the side of the road, a lot of traffic circles, etc. I keep reading online how people in SF are doing entire drives without intervention and how pleasant their experiences have been.

I find myself afraid / embarrassed to even show FSD Beta to friends and family because I do not come even close to a smooth ride while its activated, often times needing to intervene when the car completely gets it wrong.

I have a model 3, and I would like to upgrade to the Y sometime soon. I am Tesla and Elon's biggest fan but I do not think I would buy FSD again (until it improved 20fold).

Curious to get other opinions on this.
 
I find myself afraid / embarrassed to even show FSD Beta to friends and family because I do not come even close to a smooth ride while its activated, often times needing to intervene when the car completely gets it wrong.

I have had the same experience. Granted this was several months ago and I know FSD Beta has had several updates since then, but I took my friends on a FSD beta drive and it was really bad. I found myself apologizing and making excuses for FSD beta. I was like "I am so sorry. It's beta. It's a work in progress. I am sure it will get better."
 
I could have written the same words you did.

I don't think not being sold on FSD (beta) necessarily contradicts with my general favorable opinion of Tesla. I never really expected FSD to become reality on the timelines that Elon has communicated. Initially I was okay with that, figuring it was just Elon being his usual aggressive self setting stretch goals for the team. But yes, his continued doubling down on statements related to the progress of FSD compared to my own experience which barely moves the needle in terms of recognizable improvements in the 8 months or so that I've had FSD beta does frustrate me. Even more frustrating is that now he seems to have a new pet project to consume valuable AI development resources: TeslaBot / Optimus. Hey, how about we work on some of those long tail issues with FSD first?

Based on the progress I've seen so far, my expectations for when we see actual level 4 autonomy are pushed way out to the right. I'm at a point where I just ignore anything Elon has to say any more about FSD.

I'm not saying that it's not amazing what it can do today, but it falls far short of being actually usable.
 
You do find the shortcomings of regular AP when you’ve had the beta. Things like moving over for parked cars and pedestrians, ect is stuff the beta does great, regular AP stops stops for obstacles. I would take the beta over regular AP for sure. Yeah, sometimes it sucks, but when you know when it’s going to suck, you can react accordingly.
 
  • Like
Reactions: Silicon Desert
Seattle should be a lot easier than SF in general, but I'd consider SF to a bit of a cheat. Why is it a cheat? Even Elon has acknowledged that it gets much attention from Engineers along with other areas in Silicon Valley.

Even where I live it's frustrating to use.

It still doesn't properly handle 4 way stops. Sometimes it slows to soon, and it typically doesn't accelerate after stopping like it should. I'm also having issues with it turning on the turn signal in the middle of the intersection while going straight.

It also will turn on the turn signal in some corners which is embarrassing.
 
  • Like
Reactions: dhrivnak
i live in SF suburbs and honestly it's still not great. I also don't get how some people have so many intervention free drives. I have a ton of them. 12.2 phantom brakes a lot in the middle of intersections and I think this is probably one of the worst updates where i'm always pushing the accelerator for it to commit to making a move. We do have wider lanes in SF Bay Area suburbs, but if the lanes are too wide, it still has issues with placement in the lane and jerks all over the place.
 
I could have written the same words you did.

I don't think not being sold on FSD (beta) necessarily contradicts with my general favorable opinion of Tesla. I never really expected FSD to become reality on the timelines that Elon has communicated. Initially I was okay with that, figuring it was just Elon being his usual aggressive self setting stretch goals for the team. But yes, his continued doubling down on statements related to the progress of FSD compared to my own experience which barely moves the needle in terms of recognizable improvements in the 8 months or so that I've had FSD beta does frustrate me. Even more frustrating is that now he seems to have a new pet project to consume valuable AI development resources: TeslaBot / Optimus. Hey, how about we work on some of those long tail issues with FSD first?

Based on the progress I've seen so far, my expectations for when we see actual level 4 autonomy are pushed way out to the right. I'm at a point where I just ignore anything Elon has to say any more about FSD.

I'm not saying that it's not amazing what it can do today, but it falls far short of being actually usable.
just like auto-park
 

I am Not Sold On FSD Beta Experience​


Tesla did not sell you FSD Beta. What you were sold on was FSD only parts of which has been released. (I really like the FSD lane changes, though I do require my verification or request. FSD is far better and safer than I at detecting blind spot cars.)

My understanding is that the beta program is designed to collect data for improving FSD. It was not done to get reviews or to promote the eventual product.

Elon's tweets, on the other hand, are promotion of the eventual product, and so he does cherry pick for positive examples. But he is promoting FSD, not beta.

I bought FSD because I have been saying for years that machine vision will begin a huge revolution in human economies, explanation to follow. For now, FSD is the most advanced example of real world machine vision, and I want to see how it progresses.

I am sympathetic with frustration over how long it is taking, and I suspect that the delays drive much of the doubt beta owners sometimes express. But I remain optimistic that Tesla's vision and neural net approach will eventually produce safe and comfortable results, in some part due to all of our beta input.

Anyway, my advice would be not to judge FSD by today's beta. Instead, just report the glitches or opt out if it makes you feel too impatient.

About machine vision, much of what we call manual labor is actually visual labor: where to put the nails, where to sew the cloth, which strawberry to pick, which lane to drive in, etc. The industrial revolution replaced most muscle power, so machine vision will be like the second shoe dropping of the industrial revolution, enabling machines to do much of what we now call manual labor. It will vastly increase the average productive output of humans, removing the size of the "labor force" as a constraint on production of food, shelter and other necessities. As a result, the cost of these will fall. It will, of course, raise serious issues of economic equity, probably more than chain saws and cars disrupted lumberjacks and horse stall muckers, interestingly jobs which still exist. But entertainment, art, creativity, and all sorts of other human endeavors seem beyond what machines with vision will be good at. Witness the size of the credit roll on a CGI intensive film.

SW
 
  • Like
Reactions: StanMI and Hbrink
As has been said many times FSD is not ready for prime time. I find it amazing how many people have put out thousands of dollars over many years to fund its development plus agree to test it at their own risk. Unbelievable. Tesla should be funding its development and testing not the customer.
This is an example of “price discovery”.
What is price discovery and how does it work?

One difference between you and Elon Musk is that he has engaged in price discovery and has established a market for FSD beta.

You find it amazing, Elon has found a source of revenue.

You find it unbelievable, Elon has the receipts to prove it.

You suggest, without evidence, what Tesla should be doing, Elon has Tesla doing whatever they can.

As to the OP, the concepts of “caveat emptor” and “buyer’s remorse” are also applicable to “being sold on FSD beta”.
 
I'm not sure how many other products there are out there like Tesla FSD. There are many vapourware products which people do a pre-order of, but usually with a pre-order you just put down a deposit, you don't make the full payment in advance. There's things like kickstarter but they usually come from small startups, not the highest market cap companies of the world, and they don't cost this much either.

Also unusual is they way you get a small number of low-value teaser features, value a small fraction of the cost. I have not seen a kickstarter where you got something day 1, and the real product later if they pull it off.

The constant raising of the price is also unusual. As is the letting of 100,000 of the customers play with the early prototype. (That can only happen with a software product, though for early buyers this was also a hardware product as you get HW3.)

Still, I am amazed so many people buy it. I only paid $2K for it, and it's not worth that, I did it because I am a professional writer and consultant who needs to study it, not because it's useful.

Also, and this is very strange, it has almost no value in the resale market. People routinely report searching for used Teslas to buy (3rd party, as used cars from Tesla do not have FSD unless you pay the full price for it) and cars with FSD and without it do not differ greatly in price. How can that be when people are at the same time buying used cars from Tesla and paying the $12K to add this?
 
I feel compelled to add to this conversation. I did buy into FSD, although had to lobby my wife. My explanation was I can use all the help I can get as I’m not getting any younger ( currently in my early 60’s)we typically keep our cars 10 years +. I wasn’t blind to FSD’s short comings. If you let it, it will kill you (the driver doing stupid things!). I wonder what events are happening to other manufacturers ( super cruise, etc) but we don’t hear about those. I’m suspicious that’s because there are union run companies. This was clear earlier this year by our President, giving praise to union run manufacturers for taking the lead on electric cars. I get it as a political person you need to pander to unions, money and votes. I digress.

I to live in the Seattle area, Mill Creek ( north of Seattle). I use FSD as much as possible ( unless my wife gets spun up). Driving to Belfair, Arlington, Burch Bay as some examples. Has it gotten better since I picked up my MY in July, absolutely! Is it where we ( or Elon)would a like it to be, no.

The thought it’s going to get better as time goes, I’m all in on this belief. Trying to get a car to navigate on it’s own is super crazy difficult, just think of all the things that we as humans to get from point A to B. Now teach a computer to do this and all the but if this happens do this.

I’ll end on this, I saw a bumper sticker on a Tesla. It’s says “STUDENT DRIVER”, at the top it reads “AI in training “

Thanks for your time!
 
I think what’s really sad is those people that funded FSD at the beginning. Those cars are now too old and lack the sensors and computer power to use FSD. Do they get their money back ? Will Tesla upgrade their cars ? Can they transfer their FSD to a newer car ? The answer is no. Let’s not forget the more recent buyers. It’s likely that HW3 won’t be capable as well as lacking appropriate sensors. Same issues. And soon you will be paying $15K for a product that still is not capable. New York to Los Angeles any day now. Yeah right. This will play out for a few years yet. PT Barnum was right. FSD truly is Future Software Development.
 
I don’t think so. Tesla could have survived just by selling cars not software.
So says someone who joined in Aug, 2022. Have you been following Tesla since 2010 ? If not, you have no idea what you are talking about.

Afterall even in 2019 Q1, Elon was trying to sell to Apple to save Tesla from bankruptcy. Tesla got saved only because of a deal with Fiat.

There is a reason why Tesla is the first successful Auto startup in the Us in 75 years ! Its damn difficult to start a new company in such a capital intensive industry.
 
So says someone who joined in Aug, 2022. Have you been following Tesla since 2010 ? If not, you have no idea what you are talking about.

Afterall even in 2019 Q1, Elon was trying to sell to Apple to save Tesla from bankruptcy. Tesla got saved only because of a deal with Fiat.

There is a reason why Tesla is the first successful Auto startup in the Us in 75 years ! Its damn difficult to start a new company in such a capital intensive industry.
Tesla sold roughly 80k cars in 2016 and 100k cars in 2017...those are the only two years of early adopters with HW 2.0. 180k cars with an FSD take rate of ~10%. So 18k x $3k = $50M in revenue ( out of an annual $10B). I think it's safe to assume the R&D effort for FSD at the time was close to 0. (Maybe two videographers to put that fake FSD video together lol). So I guess it's not nothing!
 
  • Disagree
  • Like
Reactions: am_dmd and EVNow
Tesla sold roughly 80k cars in 2016 and 100k cars in 2017...those are the only two years of early adopters with HW 2.0. 180k cars with an FSD take rate of ~10%. So 18k x $3k = $50M in revenue ( out of an annual $10B). I think it's safe to assume the R&D effort for FSD at the time was close to 0. (Maybe two videographers to put that fake FSD video together lol). So I guess it's not nothing!
I was thinking even earlier. In any case the take rates were higher and you can see how much they collected.Just see unearned revenue item in their reports. Definitely more than 50 M.

They have 1M cars with some kind of enhanced AP/FSD out of 3M cars. So, even now the total take rate is higher.
 
I was thinking even earlier. In any case the take rates were higher and you can see how much they collected.Just see unearned revenue item in their reports. Definitely more than 50 M.

They have 1M cars with some kind of enhanced AP/FSD out of 3M cars. So, even now the total take rate is higher.

you can't assume all of that unearned revenue was from FSD. Tesla doesn't make any of that public. What we DO know is how many early adopters bought teslas in 2016 and 2017 with HW 2.0 and the option to purchase FSD. That's public. That was 180k cars.

Take rate of FSD is less of an exact science, but if you look at surveys done over the years, it was around 10% Survey Reveals Tesla's Full Self-Driving Take Rate Is Declining (insideevs.com) (you can use 12% if you want to be pedantic)

In case you forgot, tesla was still working through EAP features during that time. autosteer was barely working, there was no NoA or summon either. Much of that unrealized revenue may have been from enhanced AP