Daniel in SD
(supervised)
Has Chuck seen Tesla testing his UPL with V12 yet?
I think this will finally be the version to solve it.
I think this will finally be the version to solve it.
You can install our site as a web app on your iOS device by utilizing the Add to Home Screen feature in Safari. Please see this thread for more details on this.
Note: This feature may not be available in some browsers.
Do you think Tesla has a backlog of training data that is waiting to be used, so they'll take some trained network checkpoint to evaluate and release on a regular cycle while training the next checkpoint with the remaining data? It would seem like if it's training limited, it would be important to prioritize what data goes in next, so that would seem to prefer newer data to address situations that are still problematic with the latest version as opposed to more training of problems that have already been fixed.I would imagine we'll see FSD version updates much faster compared to v11. Like maybe an update every 2 months. Once they get into a rhythm of training and testing a new iteration as the previous iteration is downloaded to our cars, I feel like we'll see steady progress now.
Given that many employees have been driving V12 for a month, the AP team likely doesn't need Whole Mars to point out the problem areas. Tesla should have plenty of examples of driving in lots of different scenarios. Unfortunately, we only see videos from Whole Mars, who tends to post a lot of drives that are virtually identical situations.I am sure the autopilot team is watching Omar's videos. I am certain they are aware of the excess hesitancy at 4-way stops and are working on addressing it.
It's just an AI hallucination.Why does the Tesla in the thumbnail graphic have what is clearly a Waymo-like sensor pod on the roof? What's that about? I guess they are trying to make the Tesla look like a robotaxi. LOL.
What, no L5 robotaxis this year? I have already spent the millions Elon told us we would made with those…I’ve always recognizedi it will take years and I recognize that it will take years with v12 to get to a useful L2 City Streets feature that is more relaxing and safer than just driving oneself.
What makes me negative is people who underestimate the task ahead and claim we’re near autonomy. That just leads to disappointment.
What would be a reasonable miles per accident? Tesla's highest reported number was 6.57M in 22Q1, but I would think 12.x can get much higher especially given the current limited deployment/availability/usage of FSD Beta technology. I was highlighting the flattening in that Autopilot team has been busy with getting end-to-end ready almost all of 2023, so all forms of Autopilot did not see much improvement, but hopefully we'll see meaningful increases to safety across the fleet this year.it's hard to say for sure if there's a plateau or not. They may be but it's also important to remember that as the accident rate gets lower and lower it gets harder to achieve gains.
What makes me negative is people who underestimate the task ahead and claim we’re near autonomy. That just leads to disappointment.
It's only been 5 months (since end of August), but the problem could have been made worse with overweighted training to get complete stops. Potentially this is less urgent of an issue as the driver has plenty of time to push the accelerator as opposed to safety decisions that the driver might not have time to react to disengage. Specifically for the 12.1.2 hesitancy at stop signs in San Francisco, it seems like a lot of those are at hills where looking for cross traffic is quite different needing to "look" at different parts of the camera view, e.g., bottom corners of fisheye because main camera might be staring into the sky.1) 7-8 months ago, Elon's livestream showed freezing behavior at stop signs
2) Today, that behavior is still present, doesn't seem improved
Yes, but that number is for Autopilot, which is going to be dominated by highway miles. The V12 work is focused on addressing driving on secondary streets, where accident rates roughly triple. I imagine the miles per accident figure will drop significantly unless V12 has solid accident avoidance capabilities. Whether proactive or reactive.Tesla's highest reported number was 6.57M in 22Q1, but I would think 12.x can get much higher especially given the current limited deployment/availability/usage of FSD Beta technology.
Theoretically if end-to-end control is making a decision at 36 frames per second, that could be 28ms from input to output. However, average training on human examples could reflect human reaction times in initial 12.x behaviors. Although how many driving situations are truly reactionary versus anticipatory? The earlier example of potential defensive driving had context cues of adjacent lanes slowing down, so that has a lot more time to prepare, but your example of traffic light changing signals has less context except maybe for things like crosswalk countdown.Reaction time still seems to be one second
If I recall correctly Chuck has seen Tesla testers at his infamous UPL but didn't know if they were gathering data needed for V12 training or actually trying out V12. Many thought the former based on watching Tesla repeatedly taking the UPL and the consistency of each turn.Has Chuck seen Tesla testing his UPL with V12 yet?
I think this will finally be the version to solve it.
You're referring to the "left" on green from Portola to Twin Peaks (where Google Maps can't decide what's the correct left turn arrow to place on the map…)? Yeah it was quite awkward but also somewhat impressive for 12.1.2 to handle such odd signals on the fly:Impressive! The subsequent action in response to the green light was sad though. Win some, lose some.
Since it costs a lot of money to train the network once (an enormous amount of energy is required to train it), I would expect them to use whatever good training data they have and not hold any back. But that's just a guess.Do you think Tesla has a backlog of training data that is waiting to be used, so they'll take some trained network checkpoint to evaluate and release on a regular cycle while training the next checkpoint with the remaining data? It would seem like if it's training limited, it would be important to prioritize what data goes in next, so that would seem to prefer newer data to address situations that are still problematic with the latest version as opposed to more training of problems that have already been fixed.
You're referring to the "left" on green from Portola to Twin Peaks (where Google Maps can't decide what's the correct left turn arrow to place on the map…)? Yeah it was quite awkward but also somewhat impressive for 12.1.2 to handle such odd signals on the fly:
The rule isn’t difficult. The practical application is. I haven’t won a Darwin award (yet) and I still have regular instances where I stop at a 4 way stopsign and I along with the other drivers are unsure of who should go first. Who stopped first, especially when no one comes to a complete stop? What about if one driver arrives first but doesn’t actually stop but another driver arrives slightly after and does come to a complete stop - who goes first, the one who’s following the law or the one skirting it? Now throw pedestrians into the mix.Don't use Darwin award winners for training. It's not that difficult if you know the rule that the person on the left has to yield. If there's a four way tie situation then one can inch forward to judge the others reaction and safely act accordingly. There are rules, however some don't know them and others ignore them. So always yield to the unsafe driver if it calls for that.
The cars got separated so it was hard to do a true side by side comparison but in the parts I watched v12 seemed to do noticably betterWhole Mars Catalog using V12 followed by AI DRIVR using V11.
Yeah, realistically you can't get much of a better comparison. Both v11 and v12 had some tricky scenarios to deal with and both in roughly the same conditions.The cars got separated so it was hard to do a true side by side comparison but in the parts I watched v12 seemed to do noticably better