Elon tweeted that v9 of the FSD beta would remove its reliance on radar completely and instead determine decisions based purely on vision. Humans don’t have radar after all, so it seems like a logical solution and tells us Tesla is feeling much more confident in their vision AI.
Radar and vision each have their advantages, but radar has thus far been much more reliable in detecting objects and determining speed. If you’ve ever noticed your Tesla being able to detect two vehicles in front of you when you can only see the one directly ahead of you, that’s radar at work.
In this situation the radio waves from the radar sensor are bouncing underneath the car in front of you and are able to continue traveling and detect that there is another object ahead even though it could never “see” it.
It really is one of those wow moments where you can feel the future and the ability for AI-powered cars to drive better than humans one day. It’s baby steps and slowly we’ll see more and more of these situations where the vehicle simply sees or does something we could never do.
There’s no doubting that more sensors could provide a more reliable and accurate interpretation of the real world as they each have their own advantages. In an ideal world a vehicle with radar, lidar, vision, ultrasonic sensors and even audio processing would provide the best solution. However, more sensors and systems come at a price, resulting in increased vehicle cost and system complexity.
After all humans are relatively safe drivers with two “cameras” and vision alone. If Tesla can completely solve vision, they’ll easily be able to achieve superhuman driving capabilities. Teslas have eight cameras, facing in all directions. They’re able to analyze all of them concurrently and make much more accurate interpretations then we ever could in the same amount of time.
Tristan on Twitter recently had some great insight into Tesla vision AI and how they’re going to replace radar. Here’s what Tristan had to say:
"We recently got some insight into how Tesla is going to replace radar in the recent firmware updates + some nifty ML model techniques
From the binaries we can see that they've added velocity and acceleration outputs. These predictions in addition to the existing xyz outputs give much of the same information that radar traditionally provides
(distance + velocity + acceleration).
For autosteer on city streets, you need to know the velocity and acceleration of cars in all directions but radar is only pointing forward. If it's accurate enough to make a left turn, radar is probably unnecessary for the most part.
How can a neural network figure out velocity and acceleration from static images you ask?
They can't!
They've recently switched to something that appears to be styled on an Recurrent Neural Network.
Net structure is unknown (LSTM?) but they're providing the net with a queue of the 15 most recent hidden states. Seems quite a bit easier to train than normal RNNs which need to learn to encode historical data and can have issues like vanishing gradients for longer time windows.
The velocity and acceleration predictions is new, by giving the last 15 frames (~1s) of data I'd expect you can train a highly accurate net to predict velocity + acceleration based off of the learned time series.
They've already been using these queue based RNNs with the normal position nets for a few months presumably to improve stability of the predictions.
This matches with the recent public statements from Tesla about new models training on video instead of static images.
To evaluate the performance compared to radar, I bet Tesla has run some feature importance techniques on the models and radar importance has probably dropped quite a bit with the new nets. See tools like https://captum.ai for more info.
I still think that radar is going to stick around for quite a while for highway usage since the current camera performance in rain and snow isn't great.
NoA often disables in mild rain. City streets might behave better since the relative rain speed is lower.
One other nifty trick they've recently added is a task to rectify the images before feeding them into the neural nets.
This is a common in classical CV applications so surprised it only popped up in the last couple of months.
This makes a lot of sense since it means that the nets don't need to learn the lens distortion. It also likely makes it a lot easier for the nets to correlate objects across multiple cameras since the movement is now much more linear.
For more background on LSTMs (Long Short-Term Memory) see https://towardsdatascience.com/illustrated-guide-to-lstms-and-gru-s-a-step-by-step-explanation-44e9eb85bf21
They're tricky to train because they need to encode history which is fed into future runs. The more times you pass the state, the more the earlier frames is diluted hence "vanishing gradients".
Tesla’s FSD beta v9 will be a big improvement forward from what FSD beta users have been using where the system was still relying on radar. And it’ll be an even bigger leap from what non-beta testers currently have access to. We can’t wait. Now where’s that button?
Almost ready with FSD Beta V9.0. Step change improvement is massive, especially for weird corner cases & bad weather. Pure vision, no radar.
Another quarter has passed, and that means it’s time to submit questions and vote for Tesla’s Q2 2025 Earnings Call. While Q1 was a tough quarter for the company, Q2 saw some recovery in sales, although there’s still some work to be done.
However, there’s always a lot to be excited about during Tesla’s Q&A session, where we usually learn a lot about future software improvements and upcoming vehicles. We may hear more about FSD Unsupervised, Robotaxi, or the more affordable vehicle, or its upcoming larger 6-seater Model Y, the Model Y L. Tesla also mentioned a potential FSD price hike back in the Q1 2025 Earnings Call, so that could be something that is brought up as well.
Tesla’s Q2 So Far
Tesla has already released their Q2 2025 Production and Delivery numbers, which were up from Q1 of this year, but still down compared to Q2 last year.
Production
Deliveries
Model 3/Y
396,835
373,728
Model S, X, and Cybertruck
13,409
10,394
Total
410,244
384,122
How to Submit & Vote
Tesla lets shareholders submit a question that will be voted on and may be answered during the Q&A session. To submit your own question or vote on an already submitted question, you’ll need to be a verified shareholder. You can go to Say’s platform and link your brokerage accounts.
Once it is verified, you’ll be able to log in and vote your shares on your own question, or on someone else’s question.
Here’s the link to get started on Say’s Tesla Q&A. You must submit your questions and votes by July 23rd, 2025, at 4:00 PM EDT.
Top Questions So Far
Unsurprisingly, people have already been submitting questions, and here are the top ones so far.
Can you give us some insight how robotaxis have been performing so far and what rate you expect to expand in terms of vehicles, geofence, cities, and supervisors?
What are the key technical and regulatory hurdles still remaining for unsupervised FSD to be available for personal use? Timeline?
What specific factory tasks is Optimus currently performing, and what is the expected timeline for scaling production to enable external sales? How does Tesla envision Optimus contributing to revenue in the next 2–3 years?
Can you provide an update on the development and production timeline for Tesla’s more affordable models? How will these models balance cost reduction with profitability, and what impact do you expect on demand in the current economic climate?
Are there any news for HW3 users getting retrofits or upgrades? Will they get HW4 or some future version of HW5?
When do you anticipate customer vehicles to receive unsupervised FSD?
And here are some other ones we found interesting:
Have any meaningful Optimus milestones changed for this year or next and will thousands of Optimus be performing tasks in Tesla factories by year end?
Are front bumper cameras going to be necessary for unsupervised full self driving? If so, what is the companies plan to retrofit vehicles that do not have them?
Will there be a new AI day to explain the advancements the Autopilot, Optimus, and Dojo/chip teams have made over the past several years. We still do not know much about the HW4.
Earnings Call Details
Tesla will hold its earnings call on Wednesday, July 23rd, at 4:00 PM EDT. It's still early for an access link, but we’ll make sure we have a link up on the site before the earnings call that day.
If you do miss the earnings call, no worries. We will provide a full recap following the call, and we’ll also do some in-depth dives into what was said and what we know.
Tesla’s Summer Update, 2025.26, has finally launched, bringing with it a batch of interesting new features for some, and a bunch of quality-of-life improvements for everyone else.
Grok AI Assistant
The star of the Summer Update is Grok, xAI’s conversational AI assistant, which has now landed in Tesla vehicles. For now, it's available in any Tesla that has an AMD processor and is potentially coming to Intel-based vehicles in the near future. The feature is also only available in the U.S., but it’s expected to expand to other regions — hopefully soon.
Grok is in its first iteration as an in-vehicle assistant, and for now, cannot control the vehicle, which means that Tesla’s voice command system is still intact. However, there is a lot it can do already. Grok is activated by pressing and holding the voice button (right scroll wheel on older vehicles), while a short press of the button is still reserved for voice commands. Grok will support a wake word in the future, letting you activate it without pressing a button.
You don't need to sign into Grok to use it in your Tesla!
It can't currently control anything in the car, but it does seem context aware that it is in a vehicle. pic.twitter.com/IpatR7sjiJ
Once Grok is open, which can also be done by tapping the Grok app icon, users can tailor the AI personality according to their preferences by selecting a persona and voice of their choice.
There are also several other settings for Grok under the settings button. You can enable NSFW mode, Kids Mode, or disable access to your vehicle’s location.
Grok has contextual awareness of your vehicle location, which means it can provide relevant answers to questions like “Where should I go for dinner?”
Logging In Not Required
Grok is free with Premium Connectivity, or if you’re using your phone’s hotspot feature or connected to WiFi, so anyone can try it for free. In fact, you don’t even need to log in to start using Grok. However, logging in adds some additional features.
If you’d like to log in, you can do so by scanning the QR code in the vehicle, which will provide chat management and transcripts, SuperGrok access (if you pay for a subscription), and better privacy control.
Tesla has added a new Light Sync feature that pulses the vehicle’s ambient lighting in sync with the music being played. This option is turned on under Toybox > Light Sync. There are also a few options, including the ability to match the ambient light colors to the album’s artwork, instead of using your selected color.
In addition, in Park you can enable Rave Cave when parked, which cranks up the ambient lighting brightness to the maximum.
Dashcam App Update
The Dashcam app now allows you to adjust playback speeds, just like the older Dashcam Viewer, which is still used on Intel-based vehicles.
In addition to adjusting playback speed, you can now adjust the video view so that it’s displayed without being obstructed by the buttons at the top (video below). The difference is small, but could be useful if you’re trying to see something slightly out of view or that’s hidden behind the top Dashcam buttons.
While the Cybertruck has also received the updated Dashcam Viewer with this update, it does not have the new B-pillar camera recordings like other HW4 cars.
Sentry Mode is getting one of the best uses of the vehicle’s ambient lighting that we’ve seen so far. The ambient lighting will now slowly pulse red while Sentry Mode is activated to grab someone’s attention, instead of just relying on the vehicle’s display.
While you can disable Sentry Mode sounds, we’d love to see an even more stealthy Sentry Mode that also disables the ambient lighting and screen, allowing the vehicle to record without anyone being aware.
Since the ambient lighting is being used in this case to make people more aware of a feature, Tesla could also use it in other modes, such as Dog Mode.
There are a lot of potential uses for ambient lighting. Tesla can make it glow while the vehicle is charging, with the brightness potentially related to the vehicle's charge level.
When you navigate to a Supercharger, new icons in the charger list will indicate locations that require valet service or pay-to-park access.
Upon arrival at the location, a notification will appear on your screen, displaying important details such as access codes, parking restrictions, level/floor information for parking garages, and restroom availability. This information will also be available on the site card in the navigation.
Equalizer Presets
max_bracco/X
Tesla has moved the audio settings from the music player, directly into the vehicle settings — making them much easier to find. In addition to creating a new “Audio” section in settings, you can now create and save equalizer presets.
Each preset can have a name, custom EQ settings, and a setting for immersive audio.
max_bracco/X
Onboarding Guide
Not a Tesla App
Tesla has introduced a new Onboarding Guide for new owners. The guide covers driver settings, touchscreen use, steering wheel and seat setup, and how to control key portions of the vehicle, including lights, wipers, and Autopilot features.
The Onboarding Guide is automatically initiated when a new owner accepts delivery of a Tesla, or can be manually initiated at any time by going to Controls > Service > Onboarding Guide.
This appears to only be available for the new Model 3 and new Model Y.
In typical Tesla fashion, this 2025.26 update is rolling out gradually in small waves initially. Three waves have already gone out, so all signs are looking good for it going into a wide release soon.
![First Look at Tesla’s 2025.26 Update [Photos + Videos]](https://www.notateslaapp.com/img/containers/article_images/2025/grok-app.webp/f66c0424fde99ab9adb2b041d854f723/grok-app.jpg)
![Tesla Diner: First Look at Its Food and Potential Collectibles [VIDEO]](https://www.notateslaapp.com/img/containers/article_images/2025/tesla-diner-logo.webp/feaab9ffb76b08874e474c8abdf2684d/tesla-diner-logo.jpg){kind=logo}
