Tesla FSD 12.3.6 vs Waymo: Our Closest Race Yet!

Waymo and Tesla FSD Supervised V12.3.6 take on the route from McCormick Railroad Park to the Desert Botanical Gardens in Phoenix, Arizona. Waymo is an autonomous taxi that serves a large portion of Phoenix and surrounding cities.
Make sure to subscribe for more EV content!
X (Twitter): CallasEV
TikTok: CallasEV
Instagram: CallasEV
Email: [e-mail]
Webpage: www.CallasEV.com
Stickers: www.CallasEV.com/Store
Buying a Tesla? Use my referral code and earn credits for merch or SuperCharging Miles!
Calla said it's OK to use Alex's code! ;) (M3 LR Owner)
www.tesla.com/referral/alexan...
Liked Lorraine & Calla? Use her code! (M3 SR+ Owner)
www.tesla.com/referral/lorrai...
Free Music for Videos 👉 Music by Bullynho - Looking 4 - thmatc.co/?l=67EE7B9B
Free Music for Videos 👉 Music by Mak Mon, Dusty - Snow Thoughts - thmatc.co/?l=09DEEE56
Free Music for Videos 👉 Music by Mr. Chase - Check for Scratches - thmatc.co/?l=A24F101E
Free Music for Videos 👉 Music by Akkara - Sea of Clouds - thmatc.co/?l=A3088F9C
Free Music for Videos 👉 Music by Alectric - Astronomical - thmatc.co/?l=27E15218

[@thewatcher5822]

To think Tesla is doing this with just cameras is hugely impressive. With 12.4 out this week with supposedly a 5 to 10x improvement, there is only one car in this race to autonomy, and it isn't Waymo. The fact Waymo chose not to navigate the roundabout at the end shows its limitations. Tesla still has work to do, but it is taking on the much bigger challenge of driving anywhere, with the huge data lead Tesla has, and no longer being compute constrained the gap will only widen. Keep them coming Callas, These videos are great for showing the capabilities and progress of both systems.

[@justinjja2]

I wonder if it failed to navigate the roundabout or if it intentionally took that route.

[@DerekDavis213]

This was a simple 8 mile drive on a sunny day. Waymo could make this trip at night, or in the rain. No problem. Because Waymo has *much* more advanced sensors than Tesla does.

[@austinzizzi1142]

@@justinjja2 looked like waymo failed it to me

[@austinzizzi1142]

@@DerekDavis213 Tesla can also make this drive at night and in the rain clearly you haven’t been watching fsd videos

[@austinzizzi1142]

Finally someone that’s talking like they have some sense

[@Cdictator]

Please keep up with the comparative videos. I think soon there will be no more need for comparing Tesla to Waymo, because they will be always the same in certain geographical areas, but no comparison in other areas because Waymo couldn’t go there lol😂

[@ernietam6202]

thanks. It was a very good side by side comparison. Excellent!

[@lukem8420]

Yes! thank you.

[@marsec-1]

Side by side would be better. This is up by down 😊

[@nathanaelculver5308]

My only nit is that, since the Waymo was in front most of the drive, it was confusing having it on the bottom of the video; I kept thinking the upper view was the Waymo.

[@AltairIV9]

Same here hahah

[@ItsBinhRepaired]

You mean you can't tell from the interior of the cars?

[@nathanaelculver5308]

@@ItsBinhRepaired No, I don’t mean that.

[@atk0506]

@@ItsBinhRepaired Just a visual concept. The vehicle ahead feels like it should be on top.

[@ItsBinhRepaired]

@atk0506 i suppose so

[@aurora_occidentalis2248]

I use FSD every day and intervention-free drives are becoming routine. Tesla FSD is rapidly improving and will obsolete expensive over-sensored Waymo when Robotaxi becomes available. Nice vid!

[@EmanuelLangit]

Where are you using FSD every day intervention free? I live in San Francisco and during the April's free trial, I couldn't go more than a mile without intervening.

[@aurora_occidentalis2248]

@@EmanuelLangit Florida. Commute, around-town and highway driving. The latest version is much improved.

[@gomini3707]

Tesla does not take liability for any accident the FSD might cause. Waymo does for their system. Good luck finding an insurance company willing to insure Robotaxis

[@CyberCPAOH]

@@EmanuelLangitWorks pretty well in Ohio

[@assholeyeng]

with time, training and data it will get ahead

Thanks for the video; great job as always.

[@eastmanwebb5477]

Thanks for taking the time to do this comparison.

[@xh3598]

The next comparison will be in Five years because it will take Waymo that long to learn outside the restricted ZONE.

[@fredhearty1762]

With the home court advantage (pre-mapped area) and avoiding freeways, Tesla still matched the performance of Waymo. As usual, drop the two vehicles in random city -- globally -- and the result would be very different.

[@DerekDavis213]

If you drop Tesla FSD in any busy big city, without human supervision, it could not function. FSD requires 100 percent human supervision.

[@TheSkyfly203]

@@DerekDavis213 If you drop Waymo in any busy big city, without HD maps, it could not function. Waymo requires HD maps.

[@lucadellasciucca967]

@@DerekDavis213FSD is drivng in Europe and China! About to be sold in china! Your coulds and woulds are very confusing. One is a profitable business with a generalized solution and already at scale. The other a 300 car mini project, no way to scale it globally or to make it autonomous (no remote drivers supervision), losing money as we speak. Do your research, invest in Tesla.

[@thewatcher5822]

@@DerekDavis213 FSD actually could function in a busy city. There are plenty of videos of it doing just that. (without intervention) Tesla expect it to be supervised at the moment, But the car is capable of driving on busy roads.

[@landcruiser801]

@@DerekDavis213 Legal requirement versus the capability of the system... do try and keep up dear boy

[@ashegam]

awesome review and very well put together video. cheers!

[@avgjoe5969]

This is the best comparison I have seen yet. Thanks very much.

[@user-ys3in]

Thank you so much! These comparisons are very valuable information.

[@alekar2034]

Wow very cool format, hope you make more of these in the future

[@futureworld1317]

I don’t want to drive without FSD now.. This is the future…

[@U2B_Viewer]

If you time it just right sometimes without disengaging, you are able to force FSD to follow a different navigation route by using the turn signal 😊

[@lukeknowles5700]

That is a very cool and handy feature!

[@user2391-3]

this is the best waymo vs Tesla comparison video. more videos like this pls!

[@jimmyxu3819]

Awesome works

[@DamianMontero]

Love this idea. Great experiment.

[@DerekDavis213]

A simple experiment, on a sunny day. The result is misleading. People might think: FSD is just as good as Waymo. But that is *NOT* true.

[@Zedus-rl9hp]

The roundabout was a classic Waymo move: places that could cause problems are avoided. They probably do it to reduce the necessary interventions and thus improve the statistics a little.

[@Mark-kt5mh]

That's what I thought at first as well, but its navigation planner instead required it to make a u turn at a signaled intersection, which is much more likely to involve a collision.

[@jdudleyh]

Longer route = more $$

[@moreliberty1]

It still seems implausible to me that a U-turn is considered an easier or safer route than simply staying in the roundabout until the proper exit. Even disregarding the danger of U-turns, just reducing the overall length of the ride would statistically improve passenger safety.

[@Zedus-rl9hp]

@@moreliberty1 Maybe Waymo always drives like this so that it doesn't end up in the middle of that roundabout when there's a lot of traffic?! Maybe their system has problems here?

[@reinplat]

Good job on this comparison video!

[@raymondgonsalves2138]

great job!!!! excellent comparison. I would love to see a a comparison in a more challenging driving environment.

[@balaji-kartha]

There is no comparison between the two. Just because both are cars that can drive themselves does not mean they are the same! One is an animal in a zoo and the other is out in the open and free to go anywhere! One is grotesque with weird tumours all over its body and the other is a handsome beautiful elegant intelligent robot !

[@user-dn4tn9ui7w]

I am Japanese . I hope that FSD will be allowed In Japan soon. I laughed at the “weird tumors” .

[@DerekDavis213]

Yes, FSD is 'free to go anywhere' , as long as it has *100* *percent* human supervision. You are a funny guy.

[@balaji-kartha]

@@DerekDavis213 true, for now. But soon (and this could happen in less than a year) the AI would be way (like a 1000 times!) better than any human driver and then suddenly all Teslas will become autonomous! And in a couple of more years it will be illegal for humans to drive cars on public roads! Future is coming, and fast !

[@markthomas7279]

​@@DerekDavis213well yes. That's the current supervised status. You can clearly see the progression and the end game.

[@lucadellasciucca967]

@@DerekDavis213 Waymo is constantly under supervision as well. Tesla approach is simply more scalable, capable and actually a profitable business! Do your research, invest in Tesla.

[@explor794]

Nice comparison

[@xh3598]

What? How about asking for a LONG trip about 200 miles out of town? Tesla will have NO problem but the BUBLE EYE Waymo is not programmed outside the city ZONE

[@richardsauer1667]

So...Waymo has to have a "Driving Area" that is map controlled to provide it with routing capabilities, whereas Tesla can just be dropped into an area drive? This makes it seem like Tesla is "waymo" smarter than its competitor. This is not a competition. Tesla is in a league of its own. There are no competitors. Tesla Robotaxi will not need someone at a home base watching to make corrections when FSD is released.

[@flipadavis]

The Waymo vehicles also cost around $200K each while anyone can go pick up a Tesla with FSD for $50K

[@lucadellasciucca967]

@@flipadavis $30-35K

[@chicagomanny2574]

waymo smarter 🤣

[@club6525]

Mapping is not a limitation though. Mapping the entire US is not impossible. Mobileye already did it in fact.

[@jeffufcfanaticrosenberg]

wrong wrong wrong - next

[@akira28shima32]

FSD, once validated could make 3 million Tesla robot taxi. Waymo has 300 cars.

[@austinzizzi1142]

Yeah it’ll be game over real quick

[@chrispnw2547]

@@austinzizzi1142 Agree. the moment Tesla has to accept full responsibility financially for failures in the FSD software to have it approved for broad use, Elon will go silent and decline. When Musk finds the courage to take full financial responsibility for FSD errors, a possible winner could be declared and it will not be Tesla.

[@DerekDavis213]

FSD is nowhere close to being validated. Today, it cannot drive 1 hour on a challenging route without *human* *intervention* . Waymo hardware and software are much more advanced than the simple minded Level 2 FSD.

[@austinzizzi1142]

@@DerekDavis213 Waymo can’t drive an hour because that would mean it has to leave its geo fenced grid😂

[@austinzizzi1142]

@@chrispnw2547 what? Tesla will be the clear winner. People will be injured by the tech whoever’s system it is. These are all things to be navigated but just tiny things in the scheme of things. FSD will be good enough for the wider public to trust within the year. The race is over. All the billionaires are trying to throw their hats in the the ring with Wayve now which is another vision based neural net system but it’s too little too late for them.

[@lucadellasciucca967]

Definitely have the Tesla's screen visible instead of the weymo's...

[@Cdictator]

This is very impressive. Think about it, your Tesla model Y in your garage drives itself as good as a Waymo taxi…

[@CanisoGaming]

Yup, & teslas are not relying on mapped routes, + collectively learning from millions of cars driving daily. Tesla has no competition in this space, literally

[@ArielChelsau]

Certainly not true, as Tesla is L2 while Waymo is L4.

[@CanisoGaming]

@@ArielChelsau hold on to that thought while Tesla FSD keeps getting better & better every day while waymo is literally stagnant and definitely not capable enough to keep up and survive, soon enough would fade

[@ArielChelsau]

​@@CanisoGamingYeah, yeah. You Elon fanboys.

[@CanisoGaming]

@@ArielChelsau nope, my support for Tesla FSD is literally based on proven performance & clear as hell potential, I'm not investing into a stock based on me "being a fanboy" 😂

[@SanthiratanMandyam]

This is an awesome comparison!

[@juancolmenares6185]

that first left turn had me sweating! One thing I noticed looking at the Waymo from the Tesla was that the Waymo tends to stick to the right side of the lane, it wasn't completely centered. I thought that was weird.

[@davidodell6232]

I always like your versus vids. Good job with the voiceover and judicious fast-forward. I wanted to know the status of the Tesla speed control setting?

[@CallasEV]

Thanks for watching! We had the Tesla speed control set to Auto.

[@larryb8821]

Why do they show the full wheel in the Waymo and only the upper part of the wheel in the Tesla with a driver? We can't tell if the driver ever touches the wheel, right?

[@kafiluz4317]

The blue wheel icon on the top left of the screen indicates if FSD is enabled. FSD 12.3 needs nags from time to time (no steering, just showing presence). With 12.4 (rolling out this week) that nags will be removed.

[@shoelessjoe428]

This is exactly the video I was looking for. Thanks very much for your efforts. I appreciated the detailed voice over work too. I'm very split on which is better, as I see advantages for both. I prefer FSD in most ways, but it's really a gimmick until they can get rid of the 'supervised' label. A state of semi-autonomous is, in some ways, more complicated than autonomous. Because it allows for mistakes at the point of interface between human and machine. So the machine may disengage when it's not safe to do so or the human may be unsure if he should take over. Supervision can also be more stressful than driving, meaning nobody will pay for it. It has to be fully autonomous and safer in doing so.

[@leonardbrinkman4410]

And pretty soon waymo will be coming to Seattle, and that will make me very happy!

[@johnmartens5963]

Tesla FSD can operate in any country, and on highways etc

[@DerekDavis213]

FSD can operate, only with *100* *percent* human supervision. FSD cannot Self-Drive. It is Supervised.

[@brunoheggli2888]

Wrong!

[@sylvaing1]

Only in USA and Canada, for now.

[@danharold3087]

So far I have only seen it drive along a pond edge. Forget if it was dirt or gravel. I would like to see it on rural unpaved gravel roads. It eventually needs to do this.

[@danharold3087]

@@sylvaing1 Legally in the USA and Canada. We have seen it driving in Europe and China for demonstrations.

[@peterwilliams1074]

I have been looking forward to this video from you since your last FSD vs Waymo comparison. Excellent job and loved the narration. Goes to show, you don’t need to drive like a bat out of hell to get to a destination and think you will get their sooner. Sometimes, you arrive at the same time. The critical consideration for FSD is safety and I think Tesla FSD is a considerate and safe driver. Would love to see more situations that test FSD. Maybe a drop-off to the airport with a highway driving situation or going thru town with pedestrians around. Wonder if Waymo missed the turn on the round about since it was not in the correct lane. This has to do with lane selection and planning a bit better. Waymo had to do a U-turn to correct course @ 11:52. Tesla FSD was in the correct lane and made the correct turn to go straight to the destination.

[@nivvy19]

plenty of vids out there that try to push FSD

[@SteltekOne]

Waymo never intended to go around the roundabout in the inner lane (watch the map display). Maybe the Waymo Driver doesn't have enough experience with roundabouts yet, or just considers them more dangerous, so it just went for the U-turn further down the road.

[@kimwahchueng5291]

On the cuffs of new era in autonomous mobility…. Definitely coming very soon.

[@MattSeabolt]

Thanks for the content! Does WayMo usually take routes that avoid full roundabouts?

[@jeffculbert7982]

Great comparison, well done.

[@DanFrederiksen]

Curious Waymo missed navigation. It was sort of in the wrong lane but had opportunity to cross over in the roundabout. Surprising that google would fail to navigate :)

[@tripnils7535]

Chosing the inner lane is a more complicated maneuver, Waymo often choses the easier route to avoid getting in a position it doesn't know how to handle. I think this is the case here and not missed navigation. Waymo also often rather choses to turn right 3 times instead of maneuvering a complicated left turn.

[@DanFrederiksen]

@@tripnils7535 interesting. significant limitations

[@johannesdolch]

Tesla is doing this without being geofenced, without emergency remote drivers and without $150,000 in specialized equipment strapped to the roof. While this one drive might be 'close' to say that Waymo and Tesla are close in the race to autonomy is WILDLY inaccurate. Waymo's Tech isn't scalable at all, while Tesla is already at a massive global scale basically just waiting on regulators.

[@williamrogers1219]

FSD 12.3.6 can also navigate through the parking lot.

[@CanisoGaming]

Tesla has ZERO competition in this space, literally

[@richardbloemenkamp8532]

TSLA stock price says exactly what the average investor thinks about it. If I were you I'd possibly buy more TSLA stock now (depending on how much you own already of course).

[@CanisoGaming]

@@richardbloemenkamp8532 exactly, I loaded up while it was down at around 155 or so

[@Xanthopteryx]

Tesla is not self driving. Waymo is self driving. You can not really compare.

[@BTC_Minarchist]

​@@Xanthopteryxyou may be missing the point.

[@dclpgh]

​@@Xanthopteryx Take the HD maps from the Waymo and its a hunk of metal in the middle of the street. Tesla FSD does not need HD Maps. The cars figures it out on the fly with just cameras. NO LIDAR NO CONSTANT INTERNET CONNECTION, All the compute it needs is done ABOARD THE CAR. Waymo's way of doing it will not be sustainable because the sensors cost damn near as much as the car. I have an autonomous car in my driveway. And it doesn't look like a high school science project. And when Tesla throws the robotaxi switch, MY CAR will make money for me.

[@KsNoz]

This shows perfectly why you should own $TSLA stock.

[@lucadellasciucca967]

Indeed

[@brunoheggli2888]

Surw if you want to go broke!

[@overhansable]

@@brunoheggli2888 come back in 5 years and cry like a sore little loser baby

[@G41251]

Makes me thankful I bought TSLA in 2016.

[@Xanthopteryx]

Why? Tesla is not self driving (despite promised to be so almost ten years ago). Waymo is self driving.

[@Vision_Future.2024]

great vedio. more us expected.

[@joshb.1118]

For those that haven't tried it (and rely on video only) FSD made a gigantic leap. It's incredibly smooth and adept in all types of settings. And I was a skeptic. A few notes, got a new Model Y in Nov 2023 (2nd one) and Tesla provided a free month of FSD. It was good for the freeway but too choppy and abrupt for roadways (In in SoCal, Orange County). FSD went away with the initial trial at time of purchase. Then, the big FSD update happened around mid March and Tesla provided another free month of FSD. Totally diffferent experience! Smooth take offs and braking. Good decisions. Could go pretty much anywhere. I was skeptical in Nov but incredibly impressed in March-Apr. And, yes, we have roundabouts near our home, without the escape route.

[@lukeknowles5700]

The speed of Tesla's updates is awesome.

[@DuckDuckGeeseMusic]

it's amazing what 12.3.6 is doing just based on vision.

[@DerekDavis213]

But just using Vision won't last long. Tesla has recently purchased $2 million worth of Lidar sensors. Waymo has Lidar, and FSD might have Lidar too, soon enough.

[@markthomas7279]

​@@DerekDavis213I don't believe it is a comparable use of lidar. But let's see.

[@tripnils7535]

@@DerekDavis213 They didn't purchase the Lidar sensors for FSD lol, it's for training Optimus. If anything FSD 12.X shows more than ever that lidar is obsolete for driving a car.

[@lucadellasciucca967]

@@DerekDavis213 2 millions worth of LAIDAR?! 🤯 You mean enough LAIDAR for 20 vehicles?! 🤯🤯🤯 That s crazy duse. Your research is insane. Elon Musk is a about to crumble. He was always wrong 🤯🤯🤯🤯 P.S. Or maybe LAIDAR is used for 100 different things? 🤔 Nah, you re probably right

[@thewatcher5822]

@@DerekDavis213 From what I am seeing of V12, Tesla are proving that the camera only approach was the right direction. There may be some tweaks in hardware in the future, but it won't be based on Lidar.

[@DiamondMind949]

Totally unrelated, but who decided to use a right arrow as a left turn signal on that vehicle in front of the Waymo at 4:06?

[@AlternativPerspectiv]

British designers

[@Mercilessonion]

that's half a union jack

[@wlmsears]

Excellent video. It really shows the state of the art. Waymo is actually able to offer a Robo-Taxi service, which is an impressive accomplishment. On the other hand, FSD has shown that sensors are probably not the limiting factor--it's really about the intelligence to use the sensor data. Both systems work pretty well. Tesla isn't to Robo-Taxi standard, but when they get there they will have both scale and range.

[@casperhansen826]

Yet, August 8th we will know more about Robotaxi by Tesla

[@logitech4873]

​@@casperhansen826 Which of course doesn't mean anything. Tell me when it's actually available.

[@chrisja1998]

Waymo should be cited for staying in the left lane at all times. Crazy.

[@tribalypredisposed]

Like watching a race between a movie being replayed on Laserdisk and the same one being streamed on Netflix. Which one is going to win? They are so close and the movie looks almost the same…and no, Waymo Laserdisk technology is not going to beat Tesla’s instant streaming technology. The fact that Waymo is still in business incinerating billions of dollars is incomprehensible to me, they are none of them stupid enough to not realize Tesla has built the ramp to the gold ring and just has to finish climbing it now, nor can they possibly not realize that they are near the top of a local maximum dead end many miles away from the gold ring of FSD.

[@danharold3087]

To be fair anyone in self driving will be incinerating billions or they are not really in the game.

[@tribalypredisposed]

@@danharold3087 You entirely miss my point, to be fair. Tesla is INVESTING their money into their FSD program, which because they have figured out how to do it now and because they have millions of cars on the road and can manufacture millions of robo taxis in the future too will earn Tesla astonishingly huge piles of profits in the future. In contrast, Waymo believed that Tesla’s FSD strategy was impossible to do, besides their team were all experts in using LiDAR technology, so they are still working hard at perfecting what will never really work because it doesn’t scale, doesn’t work in the snow, costs way more to make a single taxi and way more to operate that taxi than a Tesla robo taxi will cost, and Waymo could never come close to having even a hundred thousand of those taxis. So Waymo is INCINERATING money, spending many billions of dollars that there is exactly a zero percent chance they will ever earn back as profits, since they will definitely never have profits.

[@ambreeshbhagtani3676]

Great video and tesla is certainly the leader here, not just in terms of scaing to any geographic location but significantly lower costs, good work tesla !

[@Alllvii]

Ive been lucky enough to get into the beta program in LA for waymo and its amazing. Love being able to just get into the car not have to deal with an angry uber driver ir a road raging one drivng me around, the first ride was a bit crazy because having a car drive itself while you sit in the back is a bit hard to grasp at first but after my 3rd ride it just became a normal thing for me. Happy am i able to witness and a ejoy a preview of what the future of driverless cars will one day be like.

[@austinterry2094]

Excellent video! I love watching these comparisons and also the unbiased narration. I believe Tesla will ultimately win in this space, and make waymo obsolete by underpricing and over performing.

[@guyguy12385]

close because of traffic and the route you picked

[@DerekDavis213]

Agreed. On a longer complicated route, or at night or in the rain, FSD would have problems. Waymo could complete the trip, with ease.

[@lucadellasciucca967]

You re right. Litterally anywhere else and FSD is the only system capable! Soon lunching in china! Do your research, invest in Tesla.

[@lucadellasciucca967]

​@@DerekDavis213 What are you talking about?! Litterally beat the Waymo at its own game! xD With no HD mapping! If it s not premapped it can t even move xD FSD has the capability if driving ANYWHERE. It s working in Europe, in China, about to lunch in China! Do your research, invest in Tesla. P.S. Waymo is supervised remotely.

[@tripnils7535]

@@DerekDavis213 FSD has zero issue with rain or night. The more complicated the route, the easier FSD would win.

[@DerekDavis213]

@@lucadellasciucca967 FSD is not capable. It requires 100 percent human supervision.

[@seanplace8192]

Honestly, I was expecting the Waymo to perform better. It works on pre-mapped road data, so it should know what each intersection looks like even before it's in sensor range. Yet Waymo chose the wrong lane in the round-about and had to make a 3-minute detour.

[@keymakerkeymaker3316]

I been using FSD for a few months now and yes I have had some really close calls but for the most part is also pretty good. With the latest update it turned on the turn signal and tried to pull in my driveway by itself, I stopped it cause that’s way too much for me lol. Also remember the FSD has various modes

[@shdmd2118]

Wow waymo car has about 200k of sensors and Tesla costs 50 k Buy Tesla stock NOW

[@DerekDavis213]

Tesla sales are down Tesla stock price is down Tesla laid off 15,000 workers FSD is years behind schedule Buy stock now is maybe *bad* *advice* .

[@shdmd2118]

@@DerekDavis213 The blind man is here!!

[@markthomas7279]

​@@DerekDavis213ah... You are revealed!! That's fine, you have an issue with Tesla. But are you an unbiased observer.... I think not!!

[@lucadellasciucca967]

Teslas cost 30-35K

[@lucadellasciucca967]

@@DerekDavis213 Tesla sales are not down. You buy when it s cheap not when it s expensive A new company is brewing Making the impossible merely late Do your research, invest in Tesla.

[@regolith1350]

Maybe it's just my OCD but it was really bugging me that the lead car (Waymo) was in the bottom half of the screen. My brain kept mixing up which car I was looking at. My brain kept expecting to see the lead car on top, and then look down at the bottom half to see the trailing car (with the lead car in view).

[@rohw0016]

Same lol

[@lukeknowles5700]

When they took two separate routes at the roundabout, there was no lead car. How did your condition handle seeing that?

[@regolith1350]

@@lukeknowles5700 It was fine because they were on separate routes, thus no impulse to do a comparison scan of passing landmarks as the lead car passed a point vs trailing car passing same point, etc. My original point was that the top/bottom lead/trailing orientation being different from what my brain expected kept messing up my ability to do a comparison evaluation during the drive. When they were on separate paths, each car was evaluated as its own thing.

[@fungilation]

more "races" like this please! Interesting divergence seen, like Waymo's failure to do roundabout

[@AltairIV9]

that's a nice video

[@TeslaElonSpaceXFan]

Tesla FSD 😍

[@xh3598]

What? How about asking for a LONG trip about 200 miles out of town? Tesla will have NO problem but the BUBLE EYE Waymo is not programmed outside the city ZONE.

[@DerekDavis213]

How about a longer drive, at night and in the rain, in the city zone? Waymo can handle it, no problem. FSD? No way.

[@spchang2000]

​@@DerekDavis213200 miles heavy rain drive in a city 😂

[@DerekDavis213]

@@spchang2000 Waymo is much more advanced that any Tesla vehicle. Waymo is in business, *today* . Tesla Robotaxi is just talk and fantasy.

[@flipadavis]

@@DerekDavis213 FSD drives fine at night and in the rain. What are you talking about? kzbin.info/www/bejne/sGKydn-OZrh5m9ksi=6zg1fOTsHJbWB-4B&t=742

[@lucadellasciucca967]

@@DerekDavis213 Your research sucks balls! xD

[@marksinclair701]

Nice comparison. I find that Waymo makes some weird pickup and dropoff choices, here in San Francisco anyway. Like looping around the block right at the beginning/end. I suspect this is for perceived safety reasons, like avoiding left turns across busy traffic, or so I don't have to cross the road. Sometimes it feels excessive (nanny state), but I presume they have good reasons. All my Waymo rides have been very impressive, Tesla may find cracking that last 10% to enable their Robotaxi quite difficult, especially if they insist on doing it without the fallback of remote supervised human input.

[@mishmash215]

I would love to see comma on here too

[@MohammedAbdullah-ir9gn]

interesting when you compare the prices of the 2 cars and how Tesla can do this anywhere in the US

[@DerekDavis213]

You are overlooking one HUGE problem with Tesla: human supervision is required at all times. Waymo drives *alone* . Waymo has much more advanced sensors than any Tesla.

[@MohammedAbdullah-ir9gn]

@@DerekDavis213 there were no interventions, they will get closer and closer to no supervision

[@DerekDavis213]

@@MohammedAbdullah-ir9gn If the Tesla guy pressed the gas pedal even once, that's an intervention. And if you watch other FSD reviews, pressing the pedal is very common because FSD often hesitates or is frozen in an intersection.

[@MohammedAbdullah-ir9gn]

@@DerekDavis213 not true, they only do that for speed

[@tripnils7535]

@@DerekDavis213 Interventions are actually super rare with the latest software 12.3.6. It's not "very common" anymore. And the next version 12.4 will improve the already excellent driving by a factor of 5-10. In a couple of months, Tesla got more driving data with V12, than Waymo got in it's entire history.

[@theodorehaskins3756]

So with FSD V12.3.6, you should’ve been able to navigate out of the parking lot at the start of this comparison test, but there’s no doubt that Tesla is the winner because unlike WAYMO that uses this expensive LiDAR system, as well as not actually being a manufacturer of their own cars, they don’t have much of a chance of actually winning the race to autonomy in my opinion. However, having said that, I am happy to see that they’re at least in the game, but they’re not competing with Tesla, because Tesla cars can drive anywhere, at highway speeds, and under all driving conditions, so they’re going after the whole enchilada, not to just be able to drive better than the average human driver, but to far exceed the capabilities of the best human driver, which isn’t even the ultimate goal, as when it comes to traffic safety, good isn’t nearly good enough. It must be as close to perfect as we can humanly possibly make it, which is driving the number of fatal crashes down to zero, and when that happens, think about the enormous value that will unleash. Cheers 🥂

[@DerekDavis213]

Tesla is not the winner. Are you kidding? Tesla FSD requires *100* *percent* human supervision. It cannot drive alone. Just ask Tesla Corporation. They will confirm that.

[@lincolntse4330]

@@DerekDavis213I would say Waymo is still ahead of the game at this moment, but their fleet is expanding really slowly, as they need to build lidar maps and all Waymo's cars are self-own while Robo taxis are owned by Tesla car owners, and there are already a few million of Tesla on the road which only require cameras. Considering the rate of improvement recently on the neural network version of fsd, I think we will know who the winner is by the end of next year.

[@theodorehaskins3756]

@@DerekDavis213 Wow! Wow! Wow! You’re trying harder to sell that line of crap than that dog I saw on KZbin that ate a block of cheese, and so nowhere has Tesla ever said that, and I noticed that you did not provide a link to where that information could be found. Hmm, I wonder why? Oh BTW! In August, a lot of people are waiting to hear about the upcoming announcement of Tesla’s launch of their Robotaxi app/car, and the unofficially named M/2 that is supposed to come in at around 25K, and so we know that Tesla requested permission, that they be allowed to test their Robotaxi service in China, and I think that they were granted that permission to do so with the stipulation that they partnered with one of the Chinese domestic car companies, and Elon said that, that company is Baidu, and that the partnership was to share the mapping data, which Tesla apparently does not possess and would have to do all of that work themselves, and so this made more, sense, for economical standpoint, and it would be much more cost-effective to just be able to tap in and share data with them, and that would also give the Chinese government confidence that they were not acting in a way that would be contrary or compromising their national security interests, which was a sticking point to allowing Tesla to self drive on China’s highways, particularly with the tension between the 2 governments continuing to ratchet up. Cheers 🥂

[@markthomas7279]

​@@DerekDavis213well of course. It's fsd supervised. But there were zero interventions. Obviously if you routed waymo out of it's mapped area it would fail immediately.

[@markthomas7279]

​@@lincolntse4330it just cannot scale. For this reason I can see no real end game. Whereas Tesla is achieving similar zero interventions in waymo areas whilst having the ability to navigate anywhere, though with the possibility of intervention being required.

[@Mojo16011973]

Loved the last couple of minutes. You should do ASMR videos also🙂

[@5thElem3nt]

Funny that the Tesla doesn't want to drive at the speed limit. The Waymo in front is constantly exact at the spped limit. The Tesla, when he is behind the Waymo falls back, driving 3 mph slower than speed limit and than comes closer again while driving faster than spped limit. Also the Tesla drives 22mph at the street to the botanical garden. This is nearly 50% over the spped limit.

[@zhli4238]

Aside from this drive result, lidar cost alone would make Waymo impractical. All cameras is the way to go. Humans use eyes like cameras. Fundamental problem is, one cannot pre-map everything, there's just too much constant changes.

[@DerekDavis213]

Human brain and human stereoscopic eyes are ultra-advanced, designed by God. You cannot compare that to Tesla cameras.

[@flipadavis]

@@DerekDavis213 Did "God" give us 360 degree vision where we can even see in near infrared?

[@markthomas7279]

​​@@DerekDavis213I think you know that's a silly argument. I am in fact religious but God has got nothing to with it.

[@lucadellasciucca967]

@@DerekDavis213 Jesus you re confused xD God doesn t exist and cameras are far better then human eyes. Which is why Teslas are better then Waymo at night, in rain or wahtever you can throw at them. Put it in Moldavia or China. GENERALIZED SOLUTION.

[@overhansable]

A tacked up 300k Jaguar is never gonna be the future of autonomy. Its the equivalent of a Nokia phone that got a screen glued on...

[@donniccolo]

Great video! Derek directed me here from a comment of one of my videos and I think it’s really funny that virtually all of the comments are in support of Tesla but he still thinks it’s terrible!

[@thewatcher5822]

Derek is simply a Tesla hater. It has nothing to do with performance.

[@donniccolo]

@@thewatcher5822 sad if true!

[@thatonedude8900]

altough the roundabout definitely is an issue here in this video and it took longer to get there but i still kinda trust waymo more for it having more sensors to base its decisions on.

[@Crawfish411-Vlog]

Let's do this test in heavy rain

[@DerekDavis213]

How about at night, while heavy rain is falling? Waymo for the win.

[@shdmd2118]

@@DerekDavis213Night and rain No difference I been using it daily. Maybe if you actually use the product you would know lol

[@DerekDavis213]

@@shdmd2118 Are you talking about FSD? It is *impaired* at night or rain, because it *lacks* *sensors* such as Lidar, Radar, or ultrasonic. Waymo has all the sensors !

[@jayan5221]

​@DerekDavis213 if there were significant problems with fsd in night/rain, then the haters would have already had a field day... vision and brain still work in night/rain...

[@shdmd2118]

@@jayan5221 yup, I use it every night, it actually think works better at night

[@JustUs-ez8ff]

Tesla can drive like waymo but waymo can’t drive like Tesla end story

[@club6525]

Compare the braking curve, though. the Tesla generally brakes and approaches cars quicker than Waymo. Tesla cannot drive like Waymo. At some point, driving becomes about those details such as safety and smoothness, and Waymo destroys Tesla in that area.

[@yaakovstern6053]

@@club6525 tesla uses software, waymo uses 100k dollar sensors. tesla is not limited to certain cities and is learning faster with faster collection of driving data. telsa is my winner

[@peterhelm522]

@@club6525Not anmyore with V12, rides are very smooth and generally without discomfort

[@peterhelm522]

Not really, when do you think Tesla doesnt need a supervisor anymore?

[@club6525]

@peterhelm522 Personally, I disagree. Yes, deceleration is smoother. But it's not human smooth. Not Waymo-level smooth. It often fails to understand priority signals unless there is a lead car like the Waymo ahead of it. It decelerates aggressively to stop. It accelerates overly aggressively off stop signs. It is smoother for sure, but its deceleration curve can not meet that of Waymo with more than 20 million real and simulated miles.

[@richardetchell1583]

Just wait until they actually start using HW4 for the input of the AI.

[@zaferroni]

Hi very good comparison its a shame Waymo is limited to a very small area .

[@elon-69-musk]

Way to go Tesla 💪

[@bearcubdaycare]

What if a homeless person gets in an autonomous taxi, and just stays and claims squatter rights?

[@pennytrui1149]

The car calls police

[@gladeous4161]

The car takes him to the cops

[@bearcubdaycare]

Will the cops do anything? Or just say it's a civil matter?

[@DerekDavis213]

@@pennytrui1149 What if the homeless guy has a cheap cell-phone jammer. Then the car cannot call anyone.

[@allangraham970]

Sure they will need to use the bathroom some time😀

[@iko359]

Cool. Love to have in EU!

[@matthewwiemken7293]

Maybe in the future we will have transponders in cars so even with limited or no visibility, they see each other:)

[@dclpgh]

There is only one Autonomous car right now. Its tesla. The rest are AUTOMATED DRIVERLESS CARS. there's a difference.

[@DerekDavis213]

Tesla is certainly not autonomous. Tesla says that 100% percent human supervision is required.

[@dclpgh]

@@DerekDavis213 then tell me what is the car doing while it's being supervised?

[@DerekDavis213]

@@dclpgh The car is being watched by a human, who will often *intervene* *or* *disengage* when FSD makes mistakes. And that happens on any challenging drive.

[@dclpgh]

@@DerekDavis213 oh boy. You're not understanding the question. What is the car doing during that supervision? Think for a moment before you answer.

[@DerekDavis213]

@@dclpgh If FSD cannot be trusted and must be *100* *percent* supervised by a human, then it is certainly NOT Self-Driving. Supervised driving is NOT Self-Driving. Now do you understand?

[@Y2Kvids]

Tesla is still at the parking lot if this is a comparison

[@kafiluz4317]

They just wanted to make sure, they are right behind the Waymo car. FSD can handle easy parking lots like this.

[@daydreamer8373]

And Waymo is still in a tiny geofenced part of America. If Tesla can solve driving across America and Canada, then Parking lots are not a big problem to solve. Tesla can already do this to a point.

[@Y2Kvids]

@@daydreamer8373 The Number of Interventions that people have to do during FSD drives would have been NEWS of accidents , if there was no driver in tesla like in waymo . Tesla has a lot to do to get to Full Self Driving and add Sensors and Lidars and Radars.

[@daydreamer8373]

@@Y2Kvids Well obviously this proves you wrong. Tesla performed excellently. Since V12 and moving to neural net, FSD has showed huge promise and is improving at a rapid pace. I am fully expecting Tesla to be operating a taxi service somewhere in America next year. And it will definitely be without Lidar.

[@Y2Kvids]

@@daydreamer8373 I doubt this , not using sensors is asking for a blind person to drive . Plus AI Decison making is a blackbox , it works when it works , when it doesn't you would have only guesswok

[@paulbentley7815]

I would choose the Waymo simply because all the spinning lidar makes it look like a vehicle from Mad Max😂😂

[@firewoodlake]

All your missing is the super mario kart course map in the bottom right corner of the screen.

[@ab22xy22]

Rigged & waymo still lost 🤪

[@ganghuangteslafsdweekly9503]

Really painful to watch the jaguar interior and screen. Please please show Tesla screen and interior next time, Tesla interior and screen is way more simplified and bring a peace of mind to the customers.

[@DerekDavis213]

Since when is a simplfied interior in a $50k Tesla vehicle a good thing? Waymo is a Jaguar. A luxury car.

[@tripnils7535]

@@DerekDavis213 A needlessly cluttered instrument panel is luxury? I think we have way different definitions of luxury.

[@PeaceChanel]

Thank You Everybody for All that you are doing for our Planet Earth.... Peace.. Shalom.. Salam.. Namaste 🙏🏻 😊 ✌ ☮ ❤ 🕊

[@Mguy393]

Nice test. I would like to see the test performed in poor weather conditions where vision based Tesla system may not perform as well.

[@thewatcher5822]

There have been many videos of FSD performing in poor weather conditions, including Snow. I have more doubts about Waymo on this front.

[@hwillia204]

Waymo uses a human remote control and map geo fencing old school. Tesla drive anywhere anytime!

[@DerekDavis213]

Tesla needs 100 percent human supervision. Tesla drives *nowhere* alone.

[@kassarc16]

Government red tape, FSD is completely capable unsupervised.

[@DerekDavis213]

@@kassarc16 You must be new here? There are about 40 FSD v12 reviews on KZbin, and FSD makes mistakes every day. Absolutely not ready to *drive* *alone* .

[@lucadellasciucca967]

@@DerekDavis213 Which doesn t mena it s not drving in china or europe, or any-fucking-where! xD Open your eyes, turn on your brain my guy

[@thewatcher5822]

@@DerekDavis213 Depends on where FSD is operating. There are many videos of FSD making mistakes, but it is driving all over America. I'm sure there are places now, where Tesla has the data to show FSD is fully capable of driving on its own right now if they chose to do it. It is only a matter of time now.

[@renehunt4252]

Waymo is old tech .It is outdated and is doomed

[@shdmd2118]

Waymo tech with LiDAR is what every company uses other than Tesla

[@markthomas7279]

​@@shdmd2118is true. Your conclusion is?? My conclusion is that FSD has a route to general autonomy whilst others do not.

[@tripnils7535]

@@shdmd2118 Yes because they don't have the advanced vision system of Tesla, others need pre mapping and lidar as a crutch. As V12 clearly shows, lidar is obsolete if you have a good vision system. Lidar is only used for localization, it doesn't make the car smarter. Localization has been solved a long time ago with cameras, it's all about the smarter driving decisions and route planning now, which lidar does nothing for.

[@lucadellasciucca967]

@@shdmd2118 Exactly the point! Nobody has taken the big risk and now they are all fucked!

[@richardbloemenkamp8532]

If all US roads are like this, then driving in the US is very easy.

[@TrendingHeadlinesTV]

FSD does not work very well on my 2018 Model S. It’s sketchy to say the least. I have to nearly every couple miles take over because FSD supervised either gets confused as to the cars around me or can’t properly see the lanes properly. Going through intersections are very scary. It stuffers and seems to lose where it’s going as I go through intersections; mind you, cars are all around me and sometimes jerks to the left or right with cars right next to me. This is so scary. The other night on the freeway, it was trying to take an exit ramp and seemed to freak out on the car behind me that was about 3 car lengths behind. FSD had the blinker on, then it slows way down, while on the freeway, then discontinued the off-ramp causing major confusion to the cars beside me and behind me. They probably thought I was intoxicated! After this scary episode, I turned fSD (Supervised) off and use normal adaptive cruise control. Way too glitchy to be safe. Also, what’s up with the 0-road speed limit at full speed? This is so scary, it launches to full power of no one is in front of you. Absolutely crazy! Tesla needs to work on a lot of glitches; especially the acceleration launch from a stop. way too aggressive and i drive on chill. There was nothing 'chill' about the numerous issues i had with FSD. I do not use it now which is a shame because it could be really an awesome piece of technology.

[@Yanquetino]

I would like to know if Waymo is pet-friendly. Could you have brought Calla with you in the car? That would be a must for me. I might be wrong, but suspect that Waymo didn't take the roundabout because the Tesla had already gotten into its blindspot, it thus couldn't move over into the turn lane, so it went straight and did a U-turn instead. As for Tesla… hold on… I thought E'loon said 5 years ago that they would have a million robotaxis on the road by 2020…!?!

[@CallasEV]

Hola! Good question, I just read their rules and no pets allowed except for service animals.

[@thewatcher5822]

Predicting the progress of FSD is not easy. Elon simply saw the progress being made and extrapolated a possible date for completion. Unfortunately FSD hit local maximums and progress slowed. The good thing about Tesla's new approach is they are not hitting any local maximums, and with FSD no longer being compute constrained, the progress will be rapid. I would not be surprised to see Tesla operating a taxi service somewhere in America next year.

[@Yanquetino]

@@CallasEV Oh… boo! Guess I won't ever use a Waymo, then.

[@Yanquetino]

@@thewatcher5822 Y'think? NOT holding my breath.

[@mrperson1324]

@@CallasEV You cant ban service animals from anywhere for the most part. They have major wide sweeping protection.

[@scooterD3]

Waymo is insanely limited. Tesla is WAY ahead of Waymo, because not only is Tesla performing a better drive than Waymo, but it can go anywhere, and doesn’t require $100,000+ added hardware equipment. Waymo will be a distant memory, since even if Waymo can get their tech to work as good, it will still be 5x as expensive.

[@christofferravn4486]

In Denmark it is illegal to stay in the left lane(s) when you are not overtaking or turning left. This does not seem to be the case in the US, seeing that Waymo is always staying out there.

[@meysamramezani3087]

So essentially the Tesla took the better route than Waymo, it works everywhere and not just in one city, is 5x cheaper to produce and is a car you can actually buy.

[@DerekDavis213]

Waymo is based on a Jaguar automobile. More luxurious than Tesla !

[@bobhope3940]

By what metric? More hand built? Yeah, no thanks. I'm positive Tesla quality and engineering is far superior.

[@markthomas7279]

That's not really the point. Autonomy is the issue.

[@DerekDavis213]

@@markthomas7279 Waymo is driverless and in business TODAY. Tesla Robotaxi is just talk and hype.

[@DerekDavis213]

@@bobhope3940 Everyone knows that Tesla has poor paint, and poor fit & finish. And a spartan interior. And no dashboard. Jaguar luxury cars have 60 proud years of history. Tesla is not even in the same league.

[@tripnils7535]

@@DerekDavis213 Seems like you need to update your knowledge. You're repeating outdated information all over the place in this comment section.

[@rb8049]

Tesla overtaking on the right almost got me into several accidents. Definitely needs fixing.

[@cellocraze]

The difference is that the Tesla is a car that anyone can buy, but the Waymo cost over 200k and it isn't for sale to the public.

[@bulolo_wilber]

Open sourcing self driving tech is the wsy to go, I hope that's part of Waymo's ambitions

[@unironicallyironic6626]

"fsd supervised" fsd stands for full self driving. *full* self driving... supervised. FULL. supervised. 😂 these people man...