John Berry
@aniccia.bsky.social
Arguments about what to call Tesla's system have moved on to the courts, for which I personally am grateful.
Uber prob books more 'robotaxi' rides than Tesla just w the 300-400 Waymo robots they handle and prob also books more rides in Teslas than Tesla. There's a metric to bother a megalomanic.
Uber prob books more 'robotaxi' rides than Tesla just w the 300-400 Waymo robots they handle and prob also books more rides in Teslas than Tesla. There's a metric to bother a megalomanic.
December 9, 2025 at 12:21 AM
Arguments about what to call Tesla's system have moved on to the courts, for which I personally am grateful.
Uber prob books more 'robotaxi' rides than Tesla just w the 300-400 Waymo robots they handle and prob also books more rides in Teslas than Tesla. There's a metric to bother a megalomanic.
Uber prob books more 'robotaxi' rides than Tesla just w the 300-400 Waymo robots they handle and prob also books more rides in Teslas than Tesla. There's a metric to bother a megalomanic.
USGS cluster, past day:
December 8, 2025 at 11:00 PM
USGS cluster, past day:
Putting new meaning in the everything app.
December 8, 2025 at 7:26 PM
Putting new meaning in the everything app.
It is a dismissal of his prob well-meaning though certainly amateurish analysis, not a dismissal of Waymo's data.
Why can you not understand this very simple difference?
Without Waymo's unreleased trove of hazard/exposure data, no one can do this calc to his precision. Amateurish overconfidence.
Why can you not understand this very simple difference?
Without Waymo's unreleased trove of hazard/exposure data, no one can do this calc to his precision. Amateurish overconfidence.
December 8, 2025 at 2:26 PM
It is a dismissal of his prob well-meaning though certainly amateurish analysis, not a dismissal of Waymo's data.
Why can you not understand this very simple difference?
Without Waymo's unreleased trove of hazard/exposure data, no one can do this calc to his precision. Amateurish overconfidence.
Why can you not understand this very simple difference?
Without Waymo's unreleased trove of hazard/exposure data, no one can do this calc to his precision. Amateurish overconfidence.
I explained right there he is not credible bc he misused Waymo's data not bc he used it.
He made the same mistake in August 2023 when he claimed Cruise's system was as safe or safer than human drivers ~2 months before they were shutdown and when there was plenty of evidence otherwise.
He made the same mistake in August 2023 when he claimed Cruise's system was as safe or safer than human drivers ~2 months before they were shutdown and when there was plenty of evidence otherwise.
December 8, 2025 at 2:18 PM
I explained right there he is not credible bc he misused Waymo's data not bc he used it.
He made the same mistake in August 2023 when he claimed Cruise's system was as safe or safer than human drivers ~2 months before they were shutdown and when there was plenty of evidence otherwise.
He made the same mistake in August 2023 when he claimed Cruise's system was as safe or safer than human drivers ~2 months before they were shutdown and when there was plenty of evidence otherwise.
Waymo has ~2k robots in California, but daily averages ~3k robots unable to move without human help.
Autonomous until stuck at an ave rate of ~10X/week per robot.
Autonomous until stuck at an ave rate of ~10X/week per robot.
December 8, 2025 at 7:13 AM
Waymo has ~2k robots in California, but daily averages ~3k robots unable to move without human help.
Autonomous until stuck at an ave rate of ~10X/week per robot.
Autonomous until stuck at an ave rate of ~10X/week per robot.
I can't speak to what antivaxxers say or write bc I have better things to do than listen to or read them, like reading the thousands of AV crash reports carefully which allowed me to identify some (not all) of GM Cruise's misleading crash reports long/months before their regulators did.
December 8, 2025 at 7:02 AM
I can't speak to what antivaxxers say or write bc I have better things to do than listen to or read them, like reading the thousands of AV crash reports carefully which allowed me to identify some (not all) of GM Cruise's misleading crash reports long/months before their regulators did.
What safety data provided by "the companies" exactly have I dismissed?
GM Cruise provided criminally misleading safety data, per GM & US DoJ.
I've never dismissed any Waymo data. I have tried to explain what it means, though.
Very few AV/ADS companies have provided "safety data" of substance.
GM Cruise provided criminally misleading safety data, per GM & US DoJ.
I've never dismissed any Waymo data. I have tried to explain what it means, though.
Very few AV/ADS companies have provided "safety data" of substance.
December 8, 2025 at 6:58 AM
What safety data provided by "the companies" exactly have I dismissed?
GM Cruise provided criminally misleading safety data, per GM & US DoJ.
I've never dismissed any Waymo data. I have tried to explain what it means, though.
Very few AV/ADS companies have provided "safety data" of substance.
GM Cruise provided criminally misleading safety data, per GM & US DoJ.
I've never dismissed any Waymo data. I have tried to explain what it means, though.
Very few AV/ADS companies have provided "safety data" of substance.
You are just making a fool of yourself w this pathetic ad hominem.
December 8, 2025 at 6:42 AM
You are just making a fool of yourself w this pathetic ad hominem.
I've posted more AV data than anyone not a regulator or mfg, inc any journal or journalist, AFAIK. I try to reference sources and be transparent. And I've been doing it for ~7 years w the same handle.
"such a narrative" is so undefined/vague and frankly childish as to be unworthy of response.
"such a narrative" is so undefined/vague and frankly childish as to be unworthy of response.
December 8, 2025 at 6:38 AM
I've posted more AV data than anyone not a regulator or mfg, inc any journal or journalist, AFAIK. I try to reference sources and be transparent. And I've been doing it for ~7 years w the same handle.
"such a narrative" is so undefined/vague and frankly childish as to be unworthy of response.
"such a narrative" is so undefined/vague and frankly childish as to be unworthy of response.
LMAO, I have explained why his analysis is flawed and should not be regarded or used. Not my fault if you haven't read it and not my responsibility to help you find it.
I referred to Prof Cummings bc she uses the term "cherry picked" to describe not just this but all of the data Waymo has released.
I referred to Prof Cummings bc she uses the term "cherry picked" to describe not just this but all of the data Waymo has released.
December 8, 2025 at 6:37 AM
LMAO, I have explained why his analysis is flawed and should not be regarded or used. Not my fault if you haven't read it and not my responsibility to help you find it.
I referred to Prof Cummings bc she uses the term "cherry picked" to describe not just this but all of the data Waymo has released.
I referred to Prof Cummings bc she uses the term "cherry picked" to describe not just this but all of the data Waymo has released.
Can't robotaxi to Vallejo's deadend turnaround atop the south tip of Telegraph Hill (pic2) bc too many robots fail there, evidently.
Waymo drops on Montgomery below steps or Kearny below the roboclot (pic1).
Glimpse the automated pure congestion of empty roboclots on the bleeding edge, on repeat.
Waymo drops on Montgomery below steps or Kearny below the roboclot (pic1).
Glimpse the automated pure congestion of empty roboclots on the bleeding edge, on repeat.
December 8, 2025 at 4:44 AM
Can't robotaxi to Vallejo's deadend turnaround atop the south tip of Telegraph Hill (pic2) bc too many robots fail there, evidently.
Waymo drops on Montgomery below steps or Kearny below the roboclot (pic1).
Glimpse the automated pure congestion of empty roboclots on the bleeding edge, on repeat.
Waymo drops on Montgomery below steps or Kearny below the roboclot (pic1).
Glimpse the automated pure congestion of empty roboclots on the bleeding edge, on repeat.
FWIW, if you do not already understand that the data he used in that opinion piece are quite cherry picked, then I suggest you consult Prof Missy Cummings, who is on linkedin.
December 7, 2025 at 8:12 PM
FWIW, if you do not already understand that the data he used in that opinion piece are quite cherry picked, then I suggest you consult Prof Missy Cummings, who is on linkedin.
You are misrepresenting my feed in fact and in spirit.
December 7, 2025 at 8:08 PM
You are misrepresenting my feed in fact and in spirit.
They may have been too busy making the final edits before US published it.
December 7, 2025 at 8:04 PM
They may have been too busy making the final edits before US published it.
More examples of other kinds of Waymo robot violations. Some of these have been happening all along, but most appear to be due to their need for better economics.
bsky.app/profile/anic...
bsky.app/profile/anic...
A robot is only an OTA away from a behavior change, good, bad, or weird.
Waymo robots have always violated some traffic laws. For >year they have been getting noticeably more aggressive as I've shown in a variety of videos. Apparently, on purpose per Ludwick who's been on this project for 15 years.
Waymo robots have always violated some traffic laws. For >year they have been getting noticeably more aggressive as I've shown in a variety of videos. Apparently, on purpose per Ludwick who's been on this project for 15 years.
“They’re bending traffic laws, getting impatient with pedestrians and embracing the idea that when it comes to city driving, politeness doesn’t pay… Waymo has been trying to make its cars “confidently assertive,” says Chris Ludwick, a senior director of product management with Waymo”
December 7, 2025 at 6:26 PM
More examples of other kinds of Waymo robot violations. Some of these have been happening all along, but most appear to be due to their need for better economics.
bsky.app/profile/anic...
bsky.app/profile/anic...
Waymo has been trying to improve efficiency, leading to an alarming increase in their dangerous & illegal driving. Many examples across diverse conditions, inc >20 recent incidents of dangerous driving around school buses & children, which NHTSA is investigating:
bsky.app/profile/anic...
bsky.app/profile/anic...
NHTSA sent Waymo a letter regarding their Austin school bus problem, strongly hinting this should be a recall which makes sense as there is no doubt it is a safety defect, ie an unreasonable risk.
I don't think ppl appreciate what a thin margin Waymo has.
static.nhtsa.gov/odi/inv/2025...
I don't think ppl appreciate what a thin margin Waymo has.
static.nhtsa.gov/odi/inv/2025...
December 7, 2025 at 6:26 PM
Waymo has been trying to improve efficiency, leading to an alarming increase in their dangerous & illegal driving. Many examples across diverse conditions, inc >20 recent incidents of dangerous driving around school buses & children, which NHTSA is investigating:
bsky.app/profile/anic...
bsky.app/profile/anic...
NHTSA sent Waymo a letter regarding their Austin school bus problem, strongly hinting this should be a recall which makes sense as there is no doubt it is a safety defect, ie an unreasonable risk.
I don't think ppl appreciate what a thin margin Waymo has.
static.nhtsa.gov/odi/inv/2025...
I don't think ppl appreciate what a thin margin Waymo has.
static.nhtsa.gov/odi/inv/2025...
December 7, 2025 at 6:07 PM
It remains unproven that there is any sustainable/reliable part on public roads without safety drivers. In California, Waymo robots fail unable to move without human assistance on ave every ~65 miles. And Waymo is almost certainly the best at this.
bsky.app/profile/anic...
bsky.app/profile/anic...
Waymo robots get stuck on ave every ~65 VMT, based on
~42 million VMT
~0.64 million stuck incidents
during the first 9 months of this year in California, as reported to CPUC for uncrewed/driverless Waymo robots in ridehail service.
Waymo's California 2025Q3 stuck ave interarrival time ~= 30 sec.
~42 million VMT
~0.64 million stuck incidents
during the first 9 months of this year in California, as reported to CPUC for uncrewed/driverless Waymo robots in ridehail service.
Waymo's California 2025Q3 stuck ave interarrival time ~= 30 sec.
December 7, 2025 at 5:35 PM
It remains unproven that there is any sustainable/reliable part on public roads without safety drivers. In California, Waymo robots fail unable to move without human assistance on ave every ~65 miles. And Waymo is almost certainly the best at this.
bsky.app/profile/anic...
bsky.app/profile/anic...
BTW, that would also explain why all the robots were empty.
Waymo's drivable map for passenger service (shown above) is more restricted than their drivable map for testing/training. Their testing map cannot be determined from publicly available sources, AFAIK.
Waymo's drivable map for passenger service (shown above) is more restricted than their drivable map for testing/training. Their testing map cannot be determined from publicly available sources, AFAIK.
December 7, 2025 at 5:26 PM
BTW, that would also explain why all the robots were empty.
Waymo's drivable map for passenger service (shown above) is more restricted than their drivable map for testing/training. Their testing map cannot be determined from publicly available sources, AFAIK.
Waymo's drivable map for passenger service (shown above) is more restricted than their drivable map for testing/training. Their testing map cannot be determined from publicly available sources, AFAIK.
Waymo had so many failures on another steep Telegraph Hill street, Greenwich, that they gave up & removed it from their robot's driveable map. This block was also removed, so it appears Waymo is still trying to master it, with difficulty.
From a thread on this w that map:
bsky.app/profile/anic...
From a thread on this w that map:
bsky.app/profile/anic...
Further evidence of Waymo's capitulation in this map of the drivable streets of and surrounding Telegraph Hill, marked green for where Waymo goes and red for Waymo no goes.
Some of these are famous addresses, eg 1360 Montgomery, and inc blocks Waymo repeatedly tried to drive and failed, then quit.
Some of these are famous addresses, eg 1360 Montgomery, and inc blocks Waymo repeatedly tried to drive and failed, then quit.
December 7, 2025 at 4:58 PM
Waymo had so many failures on another steep Telegraph Hill street, Greenwich, that they gave up & removed it from their robot's driveable map. This block was also removed, so it appears Waymo is still trying to master it, with difficulty.
From a thread on this w that map:
bsky.app/profile/anic...
From a thread on this w that map:
bsky.app/profile/anic...