Cadende

"AI" as popularly conceived (mainly LLMs) isn't necessary to automate tractors. But I appreciate the focus on productive, non-destructive uses of technology

there's some shit going on in nyc lol. I don't expect they're paying the fines. Or they're somehow making more on videos of them doing it than the fines cost.

https://www.amny.com/news/where-drivers-get-speed-camera-tickets-in-nyc/

I also can't vouch for its accuracy but I suspect its related to this: https://www.youtube.com/watch?v=JpG0LdTNfOc

this is from camera data but these are actually specific vehicles which have been caught by the cameras repeatedly: https://www.amny.com/news/where-drivers-get-speed-camera-tickets-in-nyc/

These appear to be specific cars, not just models of vehicle

thanks. I really don't get why bsky does this "the author wants you to be signed in" thing if it's just going to be publicly available via another interface.

Can you provide the link you mention? can't see any of that without a bsky account

My only qualm with this is that it seems like such blatant fraud and so easy to prove that it boggles the mind that it's flown under the radar this long.

This article actually links the lawsuit PDF:

https://driveteslacanada.ca/news/tesla-faces-class-action-lawsuit-alleging-odometer-manipulation/

https://driveteslacanada.ca/wp-content/uploads/2025/04/hinton-v-tesla-inc-et-al.pdf

Edit: uhhhhh I'm not sure I buy this. The lawsuit links this patent as evidence of tesla manipulating odometer readings: https://patents.google.com/patent/US8054038B2/en which does not say anything about the odometer, just about battery conditioning in response to predicted mileage for a given trip. Basically they integrated their charging system with the trip planner so that you can do things like plan a long route including charging stops, stop at the charger midway, and then have the car automatically charge to a minimum threshold which is enough battery to make it to the next charge point, with configurable safety margin so you don't roll in on 1% charge, and configurable driving styles so that it doesn't assume you're going to drive like a granny and therefore undershoot the amount of charge needed to go the required number of miles

If this is all they have to go on (well, this and vibes), this lawsuit is gibberish

Edit 2: they're citing reddit posts

I hate tesla as much as the next hexbear but this is not reputable stuff. The journalists that just breathlessly repeated these unsubstantiated claims because it makes a good headline should be ashamed. This took like 20 mins of research to figure out that there's no evidence provided. It could still be true but "random angry day trader sues tesla for denying his warranty claims" shouldn't be news unless there's actually any evidence of wrongdoing.

What People Are Saying

The website "isitdownrightnow" has stated: "If 4chan.org is down for us too there is nothing you can do except waiting. Probably the server is overloaded, down or unreachable because of a network problem, outage or a website maintenance is in progress."

This is journalism in 2025. Reprinting the "statement" of an obviously automated website, about the downtime of another website.

jeez I wasn't reading very carefully. I read that as "Only the RSS reader"

Only the RSS server will know the specific URL you're visiting though.

and the site itself!

I wouldn't go so far as to say it's literally the same as browsing a website. Your feed reader isn't a full web browser and as far as I know most don't execute javascript. They will still generally fetch images, and fetching the feed itself is just an http/s request, but it may or may not always be a request to the same web server as the website of whatever publication you're subscribing to. So IMO you're already starting from a somewhat better position in terms of data leakage, since the feed isn't loading analytics software or advertiser javascript or any of that stuff which feeds the vast majority of bulk data collection in the private sector.

One downside might be that if you have your feed reader set up to automatically poll for updates regularly, you may forget and it may do that polling on networks you didn't intend to (when your VPN is off or you're on school/work internet).

If you have a specific threat model, or a couple, that you want to guard against, it's much easier to come up with solutions that thwart those exact threats, than just trying to be "as private as possible" all the time (very difficult, all technical solutions have tradeoffs). You could make the requests through tor. You could use a proxy to encrypt your traffic up to a server you control before going out to the various sites. You could use a VPN service.

Those all have different tradeoffs: tor exit nodes might be widely blocked from fetching content from a lot of sites, and it might be hard to connect to tor period on some locked-down networks, the server host and their ISP can still see some details about your traffic if you run your own proxy or VPN server, but it is another step removed from your local network/isp and the site both tracking you directly by IP, user-agent, etc. VPN services might be tracking you themselves, might be working with governments, but they, similarly to proxies, interrupt the tracking done by your local network or the websites in question, with the added bonus of blending in with the traffic of other users (but they are often blocked by local network admins, and occasionally by websites as well)

As an aside, RSS-based podcasts are a place where this tends to get interesting since the field is dominated by big distribution services. Assuming HTTPS is in use, most podcasts you might subscribe to can't easily be tracked by your ISP or network admins, since they'll blend in with all the other traffic going to say, acast, libsyn, iheart, whatever, and HTTPS blocks them from seeing the full URL or data in transit, only the domain name from SNI. They can only tell that you downloaded data from a podcast network, not what podcast it was

I have been meaning to try deepseek for a chuckle/to see what the hype is about. I have pretty much no drive to use AI instead of learning or doing the work myself, but I am willing to accept that, free of the shackles of capitalism, it might be useful and non-destructive technology for some applications in the future, and maybe it's a tiny glimpse towards that