V11 Actual Experiences

[nvx1977]

This thread feels like Copywrite infringement. There already is a, “MASTER THREAD: Actual FSD Beta downloads and experience” thread that’s almost 400 pages long, so let’s create a new one and call it our own! What do you say, @Ramphex? You start the v11.3.4 experiences thread and I’ll start one for v11.4…

That thread kinda died when you became jaded. Was kinda sad to watch. I still like to look at the spreadsheet on the first post... all those people trying to get 100 safety scores.

 

[jebinc]

That thread kinda died when you became jaded. Was kinda sad to watch. I still like to look at the spreadsheet on the first post... all those people trying to get 100 safety scores.

Yeah, it’s unfortunate (having gotten jaded by my “normal” vibrating Plaid and the largely disappointing painfully slow FSDb journey). I was an enthusiastic early beta tester - one that was enrolled in the first wave back in the fall of 2021. Between the Charlatan’s never improving antics, slower than expected FSD progress and Tesla/TGC (The Great Charlatan) redefining FSD from L5 to L2, “the bloom is off the rose” .

Edit: The Master thread is still getting daily posts, but not as many as it used to. Probably because it’s now 5 months between FSD updates, whereas in the beginning it was every 2-3 weeks. For crying out loud, beta is now running five months behind the production releases. Pretty sad if you ask me. I know, you didn’t (ask me)!

 

[nvx1977]

Yeah, it’s unfortunate (having gotten jaded by my “normal” vibrating Plaid and the largely disappointing painfully slow FSDb journey). I was an enthusiastic early beta tester - one that was enrolled in the first wave back in the fall of 2021. Between the Charlatan’s never improving antics, slower than expected FSD progress and Tesla/TGC (The Great Charlatan) redefining FSD from L5 to L2, “the bloom is off the rose” .

Edit: The Master thread is still getting daily posts, but not as many as it used to. Probably because it’s now 5 months between FSD updates, whereas in the beginning it was every 2-3 weeks. For crying out loud, beta is now running five months behind the production releases. Pretty sad if you ask me. I know, you didn’t (ask me)!

Yeah it pops up here and there, but for the most part, we've moved to version-specific feedback threads (like this one). probably a better way for people to find content rather than perpetually have a single thread. But there's enough spam and crap in any thread (like this post - sorry!) that it prolly doesn't matter much.

 

[PACEMD]

Oh okay, I’m going back to the other thread then

Right, only the cool kids over here........

 

[PACEMD]

I assure you, no copyright infringement intended!

No worries, it will all come out in the hearing.......

 

[EVRider-FL]

If you click on the link "diplomat33 said" it will take you to the post.

Thanks, but why reply to that post here instead of the original thread?

 

[JB47394]

Thanks, but why reply to that post here instead of the original thread?

Because I am a forum anarchist.

That, or it was suggested that we move discussions about experienced with version 11 to this new thread. I was onboard with that, but apparently nobody else was.

 

[TicoMcC]

Lots of people have commented on V11 (in both flavors so far deployed) seeming more natural in various ways. My impression is that overall braking is quite a bit better. Mind you, it was pretty bad before and has not become remotely perfect. But I think it has slightly improved on deciding a little earlier to commence braking when the emerging situation makes it obvious to me that it should. I also think the profile of the braking during a substantial reduction fluctuates noticeably less.

I harbor a hope that my expected early service requirements for brake pads and rotors because of FSDb abuse may get postponed to later than they would have been.

I did my first drive of any real length today (62 miles, about half on interstates and half on secondary roads) with FSDb 11.3.3. My experience with braking was similar to yours, I think. My biggest issue is also that braking is still too late and too hard. I rarely use the friction brakes above 7 mph. FSDb 11.3.3 regularly has to apply the brakes around 20 mph when coming to a stop on secondary roads for stop lights and signs. BTW, I have it set to Chill mode.

So both brakes and efficiency take a hit. But yes, it is better than before.

 

[FSDtester#1]

Thanks, but why reply to that post here instead of the original thread?

Only 2 pages into the new thread and the gloves are coming off... thanks Sleepy for starting this one up. It's gonna get spicy!!
Just need more commentary from @PACEMD @jebinc @Ramphex and @Pdubs
and this place will be fire like V11!

 

[jebinc]

Only 2 pages into the new thread and the gloves are coming off... thanks Sleepy for starting this one up. It's gonna get spicy!!
Just need more commentary from @PACEMD @jebinc @Ramphex and @Pdubs
and this place will be fire like V11!

Nah, I’m out of this copycat thread! Ciao, all!

 

[Ramphex]

Nah, I’m out of this copycat thread! Ciao, all!

Samesies

 

[Pdubs]

It has been my assumption that FSD is looking at "decision points" and insists that the driver nag go out once every N decision points. I would have thought that there would be a timeout as well, but perhaps not. Certainly I see more nags when there's a lot of stuff going on around me. Merging traffic, turns, traffic lights - stuff to pay attention to.

I was noticing today on the interstate that every time I hit the accelerator to pick it up, I got the nag to torque. I put a stop watch on the nag and it was all over the place. If I was on my phone like I taught others on this board to do who shall go unnamed, I got the nag about every 15 sec. Rubbernecking got it a bit more but when I was chill and eyes front, it just didn’t nag at all.

What surprised me was the tapping the accelerator got the nag. It was repeatable.

 

[Pdubs]

Nah, I’m out of this copycat thread! Ciao, all!

I don’t even know what thread this is. I was just following my homies.

 

[tevman]

I've had 11.3 for 2 days now and done a couple of drives. General impressions:

• Overall for me it's been at least as good as 69.25 was
• There seem to be fewer phantom blinkers.
• hard to tell if lane selection is better or worse. I had it get some lanes right that it's never been able to handle before but it also got some lanes wrong that 69.25 got right.
• Turns, acceleration and stopping are smoother.
• 69.25 would often start to accelerate at a stop sign or light and then hesitate. 11.3 doesn't seem to do that.
• It doesn't intermittently hesitate at blinking yellow lights on freeway on ramps
• At the one blinking yellow arrow I've encountered it treated it like a stop sign - stopped then proceeded. Technically it should have proceeded through without stopping but it's an improvement from before.
• I've also noticed the improved creeping behavior others have commented on.

FWIW I have the voice annotation feature

First drive this morning to appointment about 40min away, residential and highways. agree with your report, somewhat hesitant at certain intersections but handled 2 roundabouts just fine, no stopping at entry. one wrong lane selection where ego couldn’t see painted arrow and there was no signage. overall very big improvement.

 

[Gwgan]

3.3 test of angled turn off Rt 1. It did make the turn, both southward off route one and northward onto route one, but both times did it badly. Going south there was an abrupt stop in traffic, a long indecisive pause but then eventually proceeded in the proper unmarked kane. Coming off the confusing road with the stop signs onto Rt 1 it did a much better job of staying to the right compared to last time but still crept onto the highway before actually taking the right turn.

 

[derotam]

Nah, I’m out of this copycat thread! Ciao, all!

Samesies

Ohh you two will be back...you can't stay away!

 

[sleepydoc]

Took a longer drive in rush hour yesterday with mixed results. on the way home it missed a cloverleaf exit. In the car's defense this is a poorly designed exit. There's a very short space in which to make the exit forcing you to watch for the traffic entering the highway before darting over 1 ½ lanes.

Getting on the next highway there was a semi trying to get on the cloverleaf I was getting off of and we were exactly in the middle of the truck. I ended up taking over and coming to close to a dead stop so the semi could make the exit. I probably could have floored it, too but it wasn’t clear. Either way it was a tricky situation.

Later on I was in a right turn yield lane turning on to a highway with traffic flowing at 55 MPH. There were 2 cars in front of me stopped, waiting for an opening. After a time FSD decided they were parked and it should try to go around them. I’m not sure how or where it expected to safely go but I didn’t let it!

Finally it messed up a double turn lane, trying to move from one lane to another mid-turn. There were no lines painted on the road but if you’re in the right left turn lane then you don’t turn into the closest lane.

On the plus side it seems like Tesla is finally using the attention-tracking software to decrease the number of nags. On my way home from work today it went anywhere from 25 - 85 seconds between nags and most intervals were over a minute.

 

[dgatwood]

I upgraded on Sunday. This is from two one-hour drives, a couple of 10-minute drives, and several drives of less than a mile, so there's not a lot of data yet from my side. But here's my first impression:

Good

• Car feels more aggressive at getting up to stop signs instead of crawling up to them.
• Turns feel smoother on average, but this may be sampling bias.

Meh

• Car isn't making me hold down the gas pedal quite as continuously on CA-17, but it is still pretty bad, and way worse than pre-FSDb using the legacy highway autopilot stack.

Bad

• Car has tried to run two red lights so far, both on left turns, both with red lights in the straight direction as well. Severe safety regression.
• Car failed to slow down for merging traffic from freeway onramps on at least two occasions, prompting me to hit the brakes or tell it to change lanes. Severe safety regression.
• The one time I made the mistake of telling it to change lanes to accommodate a merging car on the freeway, it crept so slowly that it would have likely hit the merging car had I not disengaged with the steering wheel. (Probably not a regression.)
• Parking lot navigation frequently seems to ignore the planned route, failing to make turns, turning the opposite direction from the planned route, and generally being a total mess. To be fair, I had not tried it in either of those parking lots, so this may not be a regression.

Other behavior seems largely unchanged, for better or worse.

 

[jebinc]

I upgraded on Sunday. This is from two one-hour drives, a couple of 10-minute drives, and several drives of less than a mile, so there's not a lot of data yet from my side. But here's my first impression:

Good

• Car feels more aggressive at getting up to stop signs instead of crawling up to them.
• Turns feel smoother on average, but this may be sampling bias.

Meh

• Car isn't making me hold down the gas pedal quite as continuously on CA-17, but it is still pretty bad, and way worse than pre-FSDb using the legacy highway autopilot stack.

Bad

• Car has tried to run two red lights so far, both on left turns, both with red lights in the straight direction as well. Severe safety regression.
• Car failed to slow down for merging traffic from freeway onramps on at least two occasions, prompting me to hit the brakes or tell it to change lanes. Severe safety regression.
• The one time I made the mistake of telling it to change lanes to accommodate a merging car on the freeway, it crept so slowly that it would have likely hit the merging car had I not disengaged with the steering wheel. (Probably not a regression.)
• Parking lot navigation frequently seems to ignore the planned route, failing to make turns, turning the opposite direction from the planned route, and generally being a total mess. To be fair, I had not tried it in either of those parking lots, so this may not be a regression.

Other behavior seems largely unchanged, for better or worse.

And to think, we waited a year for this (v11)…

 

[JB47394]

An FSD Uber run in San Francisco. Two minor interventions and he did the whole thing hands-free. No nags.