The reason you don't perceive as much difference as expected from the things like maps and LIDAR in the Waymo is that the difference is in reliability. They are a good part of the reason why nobody is required in the front seat, while with Tesla FSD somebody is. You need to reach "bet your life" reliability, which our Teslas are not yet close to.
Waymo is tackling things very differently. The challenges facing Waymo are nowhere near the challenges Tesla are taking on. Tesla are trying to solve a much bigger problem, and because of that may seem behind. But if you look deeper, It is Tesla who are Taking huge strides forward. I'm not saying Waymo is bad. But to me Tesla are set to dominate in this space.
Finally him giving waymo the recognition it deserves. Honestly for driverless it's really impressive not perfect it still gets stuck sometimes but that rarely happens. It doesn't happen as much as cruise waymo can even identify emergency vehicles
The fact that they allow their car to operate without a human driver is a sign that Waymo is much more mature than Tesla's software. Tesla is doing great, but can you imagine any of the Tesla FSD channels doing even a 5 minute ride without a driver? I know it won't let them, but I mean imagine the chaos it would cause. I feel like nearly every FSD video I watch, there are a few interventions to avoid a problem. Lots of respect for Tesla though, doing it all without the extra 100k in sensors.
The Waymo’s handling of the situation at 26:00 was excellent. I don’t no why the guys were clamoring on about it. If a vehicle is in the intersection and the light turns red, it proceeds.
STOP CONFUSING A TESLA SELF DRIVE FOR THIS. ITS NOT AT ALL THE SAME THING. A Tesla self drive you are 100% responsible for any wrecks. In a Waymo you are simply a guest. TESLA IS NOT AUTONOMOUS. IT CANT BE COMPARED EVER.
"Can't see any difference?", the most significant difference is that Tesla has a driver and therefore any disengagement is masked by having the driver take over, so a trip with one or two disengagements, may appears almost flawless. A Waymo disengagement is stop and call for assistance, which in itself is way more sophisticated than Tesla's FSD because there is no remote assistance with FSD. Waymo does not yet use the freeway, FSD fans would say Tesla does and therefore it is better, but Tesla does because it has a driver ready to take over. Travelling at speed, a Waymo that needs assistance hasn't got time to contact a remote operator and get guidance, and stopping or pulling over on a freeway is dangerous. Tesla may be able to drive on the freeway, but will have exactly the same restrictions as Waymo if there was no driver. Tesla will not be able to activate FSD in a car, put a taxi sign on the top and call it a RoboTaxi, there is loads more hoops to jump through before that happens. Waymo has had to learn how to interact with the emergency services, that includes the emergency services being able to geofence an area to prevent cars coming into an area, and they are trained to communicate with a Waymo remote operator and if necessary take control of a vehicle, these are all systems that Tesla will have to do before they can operate a RoboTaxi service. The Tesla FSD is a driver assist system, Waymo is a driverless Taxi service. The auto-drive systems can be compared, but not RoboTaxi feature, Tesla does not have a Robotaxi system to put up against Waymo.
Great Video. It’s now a question of scale, and we all now in time the sensor sweep will become cost effective. FSD could not handle this, great video fellas, this definitely opens my eyes to question my allotted TSLA investment. Thank you!
Hahahaha your TSLA investment????? You're really questioning your investment? Dude just by your statement it shows me you're not an intelligent investor. Do yourself a favor and just sell it. Only people who understand Tesla and the future should have the stock.
Yep … It will be interesting to see how it scales. When the novelty of these things wears off, they are going to have to compete with Uber and Yellows on a dollars-and-cents basis. My own opinion is that the lack of driver will more than outweigh the cost of the hardware hanging off these cars. And for the passenger, there won’t be some guy with their hand out expecting a 30% tip …
Tesla high share price is based on FSD potential. Waymo is not the only competitor. For instance, Wayve self driving technology seems to be as impressive as Waymo. Tesla does not have a moat with self driving car and new driving tech will emerged that leverage on multimodal AI.
Yeah, but you’re not a fanboy. A lot of Tesla groupies tend to belittle and diminish the accomplishments of Waymo and even Cruise. It’s hard not to appreciate how well these Waymos are performing if one is honest about it.
@@WholeMars - At this stage, the Tesla makes more errors. And if Tesla could handle this, they’d be running around driverless in some city. Tesla has chosen a different developmental path. As much as I’m a Musk supporter, I must admit that this Waymo is impressive.
@@Clyde-2055 FSD makes more errors because Tesla are trying to solve a much bigger problem. Tesla is developing a cheap cost effective solution able to drive anywhere. Waymo is operating in tiny bits of America. To compare the two is unfair. You say Tesla would be driverless in some city. Tesla are trying to be driverless in all Cities. That is the difference.
They're going into production this year, but it'll be a few years before they're widely deployed. This Jaguar version of the Waymo vehicle first went into production in 2019-2020 and they're only being deployed widely now.
Better is subjective. I have seen many videos of FSD outperforming Waymo in head to head comparisons in Chandler, and that was before V12. Waymo is good, I'm not knocking it. But better reaching.
Let's face it, all the videos I've seen (and I've seen a lot) of FSD are nowhere near as good as this. Tesla is probably prepping to announce they will start doing some type of premapping. In fact, they may already been doing it by using their cameras. Mark my words.
Tesla are trying to solve a much bigger problem. So while it may look like Waymo is ahead, digging deeper you will recognise the huge advantages Tesla have going forward. FSD is not ready, but when it is Tesla are set to dominate in this space.
This is really interesting -- I don't get nearly as many vocal announcements in my Waymo rides. Wonder if it's an A/B test. I've spent enough time in them though that I can just relax and don't really care what it's doing though so I'm not sure I'd want them. Also, there is an option to show the full map on the in-car screens (but I find the live traffic view more entertaining)
I have wanted for a while to see (and perhaps sort of...feel) this experience without putting my life on the line. Very interesting. Aldous Huxley would call it "Transporting-- In this Year of Our Ford."
I still contend that on board memory would help FSD so much. Rather than high definition mapping, just commit Road details to local memory so it doesn't make the same mistakes over and over again. We remember awkward intersections and blind curves, why shouldn't the car?
Because the point is to not make the mistake at all. Not make it once then remember not to make it again. With HD maps, the entire fleet knows not to make that mistake. Waymos see a construction zone, it contacts rider support, rider support determines that zone is impassible and immediately EVERY Waymo knows to route around it. Or they can help the car route through the construction zone and now every Waymo knows how to route through it.
@agildehaus Regardless, FSD has Alzheimer's. It sees every curve, every intersection over and over again for the first time. I am convinced FSD is on the right track, but it must have memory to be an AI anywhere close to human capabilities. Wamo is essentially doing this by using HD maps and human intervention.
@oscarmoreno7774 No! Not at all. I don't have a "premap" in my mind. However, I do remember preferred routes. I do remember potholes in blind curves. I do remember commonly driven roads, intersections, and obstacles. I remember how fast to take curves and when to change lanes for turns. The list goes on. Like I said, FSD drives like a person with Alzheimer's. They can do it, but because they have no memory, they make the same mistakes over and over.
Thx guys for sharing. I take it that fsd could do with better map data. I know Tesla would try and avoid being reliant on maps but there is no way past it. Poor map data equals bad fsd.
Maps aren’t necessary, helpful yes but not necessary. Picture yourself in a corn maze without a map of the corn maze. You can still find the exit but it will take time. If you had a map you would solve it instantly
@@evyb3590 I don't know. You need route planning and the better the map data available the more foresight fsd will have. FSD is the here and now (occupancy network) where the planner should be where and how to get too your destination. (which lane is the best, nearing runoff and so on)
@@evyb3590 Maps are required for a safe and comfortable experience. Vision alone is not good enough, you have to KNOW there is a speed bump there because computer vision simply isn't good enough to RELIABLY tell. Even in this video, Omar mentions FSD beta constantly missing a stop sign that the Waymo had zero trouble with.
@@agildehaus id argue that when i am driving without a map i can visually see speed bumps and pot holes to stop or avoid them in time. i also can stop at stop signs i cant see because i imply that there is an intersection that may have some traffic control and i respond accordingly until i see the control. what solves these problems is better resolution because right now the human eye has better resolution than the cameras use.
@@mariusmeyer14 ya without a doubt for route planning it is a massive help. i meant theoretically you could get to a destination given GPS co-ordinates without a map, you would just make many wrong turns and detours along the way
LOL, Chuck has PTSD on the Tesla nag. I can’t wait for Tesla to dump the nag! Omar is right, it is less safe to make the driver interact with the steering wheel while FSD is in control.
You brought up a good point with the steering wheel, I can bet you anything when this tech is fully developed there wont be a steering wheel at all, the car will be fully level 5 autonomous, still maybe 10 years or so for that I think. Very impressive though!
They have something. But they have not got it. Self driving encompasses far more than what is shown in this video. weather being an obvious one. Waymo have not solved self driving.
25:53 - not ideal to get into that situation, but that was a good way to resolve it. That's driving: an unpredictable environment that can place you in non-optimal situations and you have to do unique things to correct.
11:55 Ohhhh I would be SO jazzed to ride a Waymo down the crooked part of Lombard, assuming it's allowed (by both traffic law and the car's programming)
I'm from Newfoundland and Uber is not even allowed. We are a century away from driverless cars. Must be nice to live in a modern city that's making progress.
Hey Chuck - do we get drone footage of Omar's defeat device? Can't wait for commentary when Omar has a nag free drive after your demonstration of nags absolutely still happening. Perhaps a talking point during your drive?
It may seem that way, but in truth Tesla have some huge advantages over the competition. I have been following self driving for years, and Waymo does indeed have some impressive tech. But Tesla are trying to solve a much bigger problem, and from what I am seeing of V12 they are very close to solving self driving. From what I am seeing, Tesla are set to dominate in this space.
That was very impressive by Waymo,... much more than I expected. It would be quite interesting to see a Tesla "recreate" this scenario,... "call" an FSD Tesla from where the Waymo was summoned (the only part the Waymo really ballsed up, not shown in this cut), then have it do the Union Square (ish) to Coit Tower and back route. Also, do it on an "equivalent" day,.. Weekday/Saturday/Sunday, same time of day, and with similar weather conditions. I have significant doubts that the Tesla FSD will be anywhere near as polished. I'm getting increasingly concerned about FSD achieving level 5 by 2025 which was my expectation
With all of Elon’s talk about robotaxis, and given the progress Tesla has made with their FSD product, I’m a bit surprised he hasn’t made the move to set up a driverless ride-share service somewhere like Waymo is doing. While Tesla’s developmental model differs, they still are targeting a robotaxi service.
I think human drivers are less patient with Tesla when driving in FSD. It’s like they expect the Waymo or Cruise to act like a robot driver but Tesla acts that way and people honk at it.
Maybe because these cars actually look like robots you can't really see if a Tesla is using fsd. While this car has like five thousand sensors everywhere even a big large one on top that constantly keeps spinning. But people have also been honking at waymo if you look at videos
19:50 It did not "do the right thing", that's a violation of CVC 22526 as soon as the car enters the intersection without room to clear, "Blocking Intersection (Gridlock) Prohibited," $281 fine
The cost of Waymo with all the hardware compared to a Tesla just using 8 cameras is no contest. There are now 5m Tesla vehicles on the road with the necessary harware to turn into Robotaxis with an overnight software update, so ready tomorrow! So slight advantage on cost and scalability 🤔
Well waymo uses way more cameras because if it crashes the company is at fault or can be sued. But Tesla can't be held liable they will just say the driver should have been paying attention. Waymo is more about safety which is why they don't even go on hightways yet i think their scared to get sued if the car makes a mistake 😂 Edit: this car is definitely better then a tesla at the moment literally like 2 videos ago he had to correct the fsd or it would crash. No one is behind waymo if something suddenly happens it makes all decisions itself. Only time people take control is if the car gets stuck
@@pennytrui1149 Waymo is not definitely better than FSD. In isolation it may seem that way, but taking a step back and seeing the bigger picture it is Tesla holding all the aces. Waymo does have some impressive tech, and that deserves recognition. But so does Tesla, it is just not as obvious.
Comparing Tesla to Waymo is largely inappropriate. The objectives of each company are different at this time. Waymo's objective is operating a tightly constrained Level 4 robotaxi service. Tesla's objective is to crack the core fundamentals of generalized self driving at scale. Tesla could easily do what Waymo is doing if they chose to put similar constraints on vehicle operation (eliminate difficult unprotected left turns, avoid highways, provide a remote driver, etc). Waymo cannot do what Tesla is doing without completely architecturing their system.
Well waymo doesn't do it because any crash that the car makes makes the company liable or can they be sued. Tesla doesn't care about that because the fault is always the drivers fault waymo will go on highways when it's safe.
@@LeoMoralesSr Which is why Tesla are set to dominate in this space. Waymo as impressive as it may seem has some huge disadvantages compared to Tesla, Who have now accrued billions of miles of real world data, essential to solving self driving. Tesla have always said their system is level two, anyone following it's progress will know it is far more capable than that. It is not ready, but from what I am seeing of V12, they are getting very close to solving self driving.
Lidar gives an absolute distance value so that's easier to work with but vision can do it too. I saw it get stuck in a 3 point turn in another video. It just lost its nerve, plenty of room and then the remote spirit possessed it and made it better. And yeah I have hopes that taxipods will kill off old public transport as the price can be so low that a bus will never be interesting. The success level can actually become a problem. We'll need boring tunnels to boost road capacity. Or into the air.
