Not to mention there isn’t public pushback against an office with a server room like there is against dedicated data centers. Assuming they are running servers and not just some desktops with GPUs.

Honestly, I’d suggest first finding which cars have guides to remove or disable that shit and go from there.

I’ve been impressed with how it handles multimonitors at different resolutions. It scales it dynamically so that (for side by side) the top of the one screen lines up with the top of the other and same with the bottom, plus windows look like they are the same size (relative to screen size) smoothly even as they overlap both monitors.

The only thing I don’t like is that the cursor can get “stuck” at the edge of one monitor, though I think that behaviour might be gone now after a recent update because I haven’t had to fight it as much lately.

As someone who assumed that it wasn’t done like that because the scaling would look bad because pixels would need to be scaled to subpixels, I’m pretty impressed that that doesn’t seem to be the case (either that or they figured out how to handle that scaling so that it didn’t look like shit, though this might be the very thing that the custom renderer is failing to do properly in your case).

Not sure if this is still the case, but I see some do the reverse to get around the “no extra fees for using a CC” rule: instead of adding fees when using a CC, they set the prices higher by that fee all the time but then offer a discount for cash or debit payments.

Yeah, that and I was already wary of Google by the time using your phone to pay was a thing and it was an even bigger “fuck no” than when they and MS wanted me to hand over my files for them to store instead of using the perfectly good storage I have at home.

Edit: replied to wrong comment

Hell, even the fact that the term “African American” was invented because “black” was considered offensive. Judgement was baked right in (or at least assumed to be by enough that this somehow got popular) to using their skin colour as a descriptor.

Which then led to many people being called “African American” despite having no connection to either Africa or America.

This is another one of the windows negatives. There’s no real standard for tools that debloat or remove the MS malware and decent chance that you end up getting more malware in the process (that might even do what it says it does in addition to installing who knows what, so it’s not even immediately obvious that it might be a good idea to go back to the “wipe storage and install OS” step. Thing could even revert some of the settings later on to encourage users running an updated version of the script (not that I think most would bat an eye at their “OS optimization script” being a thing that continuously runs and auto-updates) and it wouldn’t even be unexpected windows behavior considering how passively hostile they already are with users trying to reduce their stupid shit in windows.

Yeah, personally, I’ve noticed that I notice and appreciate very high quality streams when they are there but don’t notice lower quality ones in a bad way (where “lower quality” is still like 1080p, 720p is more noticeable).

Like 4k looks great but 1080p still looks normal.

Or just do it like reddit did, where you can delete your post content and remove your username from it, but the thread and comments remain.

Though with how the fediverse works, it’s possible to spin up a custom instance that highlights deleted content instead of deleting it, meaning the attempt to get rid of it can be what brings it more attention if anyone has decided to do it. Just like with vote identities, they aren’t anonymous and there are instances/sites that just show who voted for what.

I’m still undecided about their own text (deleting comments or post text disrupts the conversation but there’s valid reasons), but I don’t think the OP of a thread should have any control over the existence of that full comment thread.

Frankly, I don’t think users should have that power.

There was an article about removing the modem (or other essential piece of that data path) from a specific model of car (prius I think?), hoping to see more of those.

Yeah, reinventing the wheel can be fun.

Yeah, I kinda wish the site generated a hash or something because I’ve got an extension that fakes the canvas results, but the site says those identifiers are unique for me… But are they the same unique (which indicates the extension isn’t doing anything) or different each time (which might even make the others less useful if it aggregates everything?

I did notice earlier today that the YouTube recommendations were all actually related to the video I was currently watching instead of it trying to get me to go down a rabbit hole I’ve already been down even logged out, like it does on my desktop where I haven’t installed that extension.

I like that they covered all the possibilities for the do not track flag, as I saw it as useless from the very start, as by then I realized the honour system didn’t mean shit and it would just be another piece of data.

It actually sounds like they are allowing bots as long as a human can intervene and is willing to verify their identity if their automated activity is detected. They are embracing slop users as long as they give them their data.

I think the main reason for the announcement, though, was to deal with the rumours that everyone will need to verify their identity (which this post seems to assume is still the case).

Personally, I’d rather see reddit remain the mainstream version of this just so that people who want to run bots for whatever reason have more reason to do it there than here. Because the commercial level ones will have more resources than lemmy admins will have to deal with it.

Just flashbang every room before you enter it, just to be safe.

Yeah, it’s good enough that it even had me fooled, despite all my “it just correlates words” comments. It was getting to the desired result, so I was starting to think that the framework around the agentic coding AIs was able to give it enough useful context to make the correlations useful, even if it wasn’t really thinking.

But it’s really just a bunch of duct tape slapped over cracks in a leaky tank they want to put more water in. While it’s impressive how far it has come, the fundamental issues will always be there because it’s still accurate to call LLMs massive text predictors.

The people who believe LLMs have achieved AGI are either just lying to try to prolong the bubble in the hopes of actually getting it to the singularity before it pops or are revealing their own lack of expertise because they either haven’t noticed the fundamental issues or think they are minor things that can be solved because any instance can be patched.

But a) they can only be patched by people who know the correction (so the patches won’t happen in the bleeding edge until humans solve the problem they wanted AI to solve), and b) it will require an infinite number of these patches even to just cover all permutations of everything we do know.

Here’s an example I ran into, since work wants us to use AI to produce work stuff, whatever, they get to deal with the result.

But I had asked it to add some debug code to verify that a process was working by saving the in memory result of that process to a file, so I could ensure the next step was even possible to do based on the output of the first step (because the second step was failing). Get the file output and it looks fine, other than missing some whitespace, but that’s ok.

And then while debugging, it says the issue is the data for step 1 isn’t being passed to the function the calls if all. Wait, how can this be, the file looks fine? Oh when it added the debug code, it added a new code path that just calls the step 1 code (properly). Which does work for verifying step 1 on its own but not for verifying the actual code path.

The code for this task is full of examples like that, almost as if it is intelligent but it’s using the genie model of being helpful where it tries to technically follow directions while subverting expectations anywhere it isn’t specified.

Thinking about my overall task, I’m not sure using AI has saved time. It produces code that looks more like final code, but adds a lot of subtle unexpected issues on the way.