Pinned
John Berry
They are in the safety walker stage (pic1), prob won't make it to the remote walker (telop) stage. These Chinese robot dogs and related should come under San Francisco's 2017 law (pic2) that is the only reason SF hasn't been inundated w sidewalk robots, or SFBOS should update the law to make it so.
A friend in SF was out walking his dog and saw someone else walking a robot dog. He sent me the video. The other guy was a paid robot dog walker. I had SO many questions so I investigated: www.nytimes.com/2026/10/27/t...
The big declines are from bad incidents, the smaller bumps up are from Waymo marketing campaigns. The downward trend reflects how the reality falls short of promises. Politicians and regulators decide whether, where, and how AVs operate and they are much more amenable to Waymo's $$$$$$ persuasion.
YouGov finds that more Americans have heard about Waymo, but the company’s reputation is net negative. “The data suggests that growing familiarity is not necessarily translating into more favourable perceptions of the brand.” yougov.com/en-us/articl...
Yes, though the key is none of this is going to happen and ppl are just wasting time dreaming about it. Robots are so bad around school buses that Waymo had to make their's stop and wait for a human to help them just drive by. Also, transit needs to be reliable, not shutdown for rain, snow, etc.
If I were trying to get people comfortable with riding buses with no adult employee on board, I'm not sure I'd start with school buses. The hand-waving here is not enough to comfort me given the obvious "Lord of the Flies" scenarios. open.substack.com/pub/changing...
This 2022 GM Cruise block of the N-Judah in Cole Valley was in SFMTA's letter to CPUC objecting to Cruise's key request to expand their operation. SFMTA wanted more regulation and reporting, which CPUC ignored at the time but ~adopted yrs later after the Cruise debacle they v much allowed to happen.
All FMVSS complaint vehicles can also be left directly unsupervised in a geofence and on average they are ~95% of the time, ie same as Waymo per their co-CEO. There's no SAE ODD language spec or similar to match their AV functional spec (J3016) and no minimum domain required in theory or practice.
I mean they can be left directly unsupervised in a geofence, Waymo has ~4k such vehicles on the road right now, but there is still remote oversight and occasional remote guidance and intervention for customer service stuf... no robot ever has had zero humans in the loop over its product life
icymi, video begins (pic1) as a Waymo robot forces 3 drivers to back up, 1 robot & 2 human. It ended (pic2) w the ~7 ft wide Waymo robot given >11 ft between a Zoox on left and parked cars on the right, but having driven toward the Zoox instead of bearing right into this space & back into its lane.
Apparently, a Waymo robot drove around a stopped ~half in lane Tesla robot to bully a Zoox robot that prob had the right of way to back up: Gibson by Congress, Austin, TX All 3 cos have telops, but did any engage? AV advocates say the best solution is more robots. OP: reddit.CautiousMagazine3591
While this video is novel bc 3 flavors of inept robots, there are many weirder videos and of more dangerous/unsafe robot driving that only involve Waymo robots struggling to pass each other on narrowish streets, eg they endangering 7 pedestrians in this one: bsky.app/profile/anic...
"cars were just holding down their horns because Waymo was holding up the street for 3 minutes. Sped up...7x...Am I the only one tired of Waymo clogging up narrow streets of San Francisco? They really need to start utilizing bigger roads than the narrow backroads." OP: tiktok.true.wrong
A better view of this Waymo behavior. Waymo upgraded their I-Pace robot microphones for sirens after SFFD complaints, but well before these incidents. This seems to have begun after/when Waymo purposely made their robots more "aggressive," same as passing school buses. bsky.app/profile/anic...
A Waymo robot impeded LA County Firetruck 172, despite sirens: Centinela Ave at La Brea Ave, Inglewood, California Not sure how much more evidence regulators or anyone else should need to realize these robots do not have sufficient situational awareness and that it is a glaring & dangerous defect.
In 2022, SF gov gave NHTSA an accounting of GM Cruise impacts, inc many good/ignored recommendations (pdf at link): - detailed AV/ADS failed in traffic lanes - illegal pickup/dropoffs - Cruise's Critical Response Line failed to answer calls from SF 911 dispatchers regmedia.co.uk/2022/09/26/l...
This was the one on Carl in September. GM Cruise also blocked N-Judah at Carl & Cole. They had to send someone to drive it away. Robots without safety drivers = blocked transit. IDK of a public accounting of Waymo's transit impacts in San Francisco or anywhere else. bsky.app/profile/anic...
IDK if this is the most transit riders blocked by one robot, but it is a contender. Safety driver fixes this in seconds, but The Waymo Driver has been too good to need them since October 2018, according to California DMV. OP: x.cammakingminds
"Bailey Koch-Farwana, a [Service Employees International Union Local 1] spokesperson, said that according to their economic analysis, every AV that hits the road replaces four human drivers. This is because AVs can operate 24/7 and do not need breaks." Only in dreams. Waymo robots ave ~12 hrs/day.
Buckner on Giannoulias’ self-driving car concerns: ‘I’m just assuming he’s been a little bit busy’
IDK much about DC's Waymo debate, but this seems reasonable. I'd suspected DC historically had an abundance of cheap taxis to keep their abundance of drunk Congressmen out of the news, but didn't know DC had 9k taxis 30 years ago. SF had ~1.5k for more ppl pre-Uber. dc.citycast.fm/opinion/dc-i...
Cortland, like 24th St in Noe Valley, has one non-parking lane each way and is the main neighborhood business street. In case anyone is wondering how one failed robot could block the whole street & buses going both ways, a Waymo robot did that on 24th St in Noe yesterday: bsky.app/profile/theo...
Beware, a rideshare driver app co dumped some (not the key) economics data in a "report" on AV impact w questionable methods/disclosure. States that permit AV ridehail and literally have ALL the receipts should prob do these studies ~annually. 7021468.fs1.hubspotusercontent-na1.net/hubfs/702146...