The author actually explains that his original solution was just saving them locally on his phone and playing them from there, but that was too much legwork for his wife to want to switch from a cloud service like Audible. So the whole self hosting part is to become “Audible” for his wife lol.
I mean if they were doing this already there would be no point in sending this email out. They would have just happily continued letting people think it wasn’t happening while doing it anyway, while not having to deal with the backlash this will generate.
Actually as much as I’d love to use Firefox all the time, there are many times it won’t work properly at all. This isn’t entirely Mozilla’s fault, but it is the case.
The problem is that the sales pitch of these AI answering services in search engines is to save you time from having to open search results and read them yourself. The problem with 80-90% accuracy is that if the summaries are hallucinated even once, you can no longer trust them implicitly, so in all cases you now have to verify what it says by opening search results and reading them yourself. It’s a convenience feature that doesn’t offer you any actual convenience.
Sure it’s impressive that they are accurate 80-90% of their time, but AI used in this context is of no actual value.
If you couldn’t suss out the obvious sarcasm, it might be time to go offline for awhile and recalibrate.
You can get a VPS on Vultr for $5 a month that’ll be plenty for your use case. For security look into mTLS. Web servers like Caddy support it pretty easy. You put a certificate on you and your wife’s phones and only you will be able to connect to your endpoint.
The definitions you’ve linked above do not contradict what I’ve posted below or how I’ve described the behavior in the article. You’re running a narrow definition, I’m running a broad definition, the distinction is really fine and neither is wrong.
From the linked Wikipedia article I linked above:
Ok so a couple things here. You mention your knowledge is decades old. If you’re talking more than two decades ago there’s a chance your definition was correct a very long time ago. But to be clear, today in the year of our lord 2025, in the United States of America, your definition is not correct. There are no “narrow” and “broad” definitions. There is a single definition for each in the context of firefighting in the U.S.
One of the things you’ll want to get in the habit of doing, is not quoting Wikipedia as your source, but rather the source that Wikipedia refers to when it says something:
there is considerable disagreement about categorizing backdrafts as flashovers
The disagreement mentioned here refers to the long history of loose definitions in the distant past of fire science as these phenomena were being studied and categorized, as well as differences between scientific and firefighting applications, and European/ US entities. If you follow that reference in your quote [5],
You’ll see it goes over the history of the terms and how we’ve ended up with the current definitions. In the U.S. based fire service, the be-all end-all of these terms in NFPA 921-2004 which is as follows:
The newly crafted NFPA 921-2004 definition of flashover is: “A transitional phase in the development of a compartment fire in which surfaces exposed to thermal radiation reach ignition temperature more or less simultaneously and fire spreads rapidly throughout the space resulting in full room involvement or total involvement of the compartment or enclosed area.”
The source goes on to explain:
However, for firefighting purposes, the NFPA have recognized for over twenty years (in their annual reports concerning firefighter life losses) that there are several other forms of related phenomena or terms used, such as smoke explosion; flameover; backdraft; flash fire; etc and that many of these phenomena cannot be explained or directly attributed by on-scene firefighters.
Therefore, the NFPA reporting system has established the term ‘Rapid Fire Progress’ to cover all situations where some form of fire phenomena led to an extreme event of combustion causing sudden transition from a small fire to a large fire, even where flaming is not sustained. They further refer to various sudden or extreme fire phenomena as falling into one of three categories:
The whole “All backdrafts are flashovers but not all flashovers are backdrafts” comes from studies done by the Swedes in the 1980s which, depending on how long ago you spoke with your lieutenant, may have been what he was referring to. The source article touches on these as well, but again, these are not acceptable terms today.
So if in your article you wanted to use the non-specific broad definition of these phenomena, you’d use “Rapid Fire Progress”. But if you wanted to use terms like flashover or backdraft, you’d need to use the correct one for the phenomena you’re trying to describe because they are not interchangeable any longer.
Here is another great article in PDF format about RFP and its categories that really goes into detail on their differences. It also explains the issue I take with your article, which is that these terms are still to this day sometimes casually misused as if they’re unimportant and interchangeable. This is harmful when presented in an educational context (like your article is to some degree), because these terms need to be clear, distinct, and have a universal meaning so that conditions can be accurately communicated when you’re fighting a fire.
Why is the article using diagonal screen size as their measurement for phone size? In that case you could have a phone the exact same size get “bigger” just because bezel sizes have shrunk over the years.
They specifically call out the iPhone SE as a “small phone” that they seem to want. But the newest iPhone, the iPhone 16 is only 6% bigger in width and height. Fractions of an inch larger. I can totally understand why somebody would want a phone with smaller overall dimensions, but why on earth would your metric for an ideal phone be a smaller screen?
Samsung, Xiaomi, Oppo/OnePlus, and Vivo, the top four Android smartphone manufacturers, have not released a single phone with a display 5.5 inches or smaller in the past three years, according to data from GSMArena.
Seems like they’re going away on the Android side too.
Flashover and backdraft are two different things with two very different mechanisms behind them. The naming of each are standardized across the nation as each academy should be teaching to the NFPA 1001 standard at a minimum.
In order to address that with your lieutenant, I would direct him to NFPA 921.
Pretty interesting, though the current primary method of EV Fire fighting tactics is the use of fire blankets, which isn’t as exciting or game worthy.
Also lol @ the author confusing backdraft and flashover.
You can use programs like Ripper or Automated Ripping Machine to watch for inserted disks and rip them automatically. You can even add remixing or transcoding into the workflow so that all you do is insert the disk into the drive and after some time out pops the file in the exact format and size you want, even directly into your media library if you want.
I’d imagine that’s the case for a lot of people actually. People don’t just pirate to get things for free. Lots of people buy easily pirateable games on GOG because it doesn’t have DRM. Lots of people buy and rip Blu-rays instead of torrenting movies.
While that’s true, the thing that stuck out to me is not even that the AI was mislead by itself finding AI slop, or even somebody falsely asserting something. I googled something with a particular yea or no answer. “Does X technology use Y protocol”. The AI came back with “Yes it does, and here’s how it uses it”, and upon visiting the reference page for that answer, it was documentation for that technology where it explained very clearly that x technology does NOT use Y protocol, and then went into detail on why it doesn’t. So even when everything lines up and the answer is clear and unambiguous, the AI can give you an entirely fabricated answer.
I was just talking about this with someone the other day. While it’s truly remarkable what AI can do, its margin for error is just too big for most if not all of the use cases companies want to use it for.
For example, I use the Hoarder app which is a site bookmarking program, and when I save any given site, it feeds the text into a local Ollama model which summarizes it, conjures up some tags, and applies the tags to it. This is useful for me, and if it generates a few extra tags that aren’t useful, it doesn’t really disrupt my workflow at all. So this is a net benefit for me, but this use case will not be earning these corps any amount of profit.
On the other end, you have Googles Gemini that now gives you an AI generated answer to your queries. The point of this is to aggregate data from several sources within the search results and return it to you, saving you the time of having to look through several search results yourself. And like 90% of the time it actually does a great job. The problem with this is the goal, which is to save you from having to check individual sources, and its reliability rate. If I google 100 things and Gemini correctly answers 99 of those things accurate abut completely hallucinates the 100th, then that means that all 100 times I have to check its sources and verify that what it said was correct. Which means I’m now back to just… you know… looking through the search results one by one like I would have anyway without the AI.
So while AI is far from useless, it can’t now and never will be able to be relied on for anything important, and that’s where the money to be made is.
Lack of feature parity is the number one thing holding so many people back from switching to Jellyfin. Of Plex is going to start deleting beloved features, a lot of minds will be made up very quick.
Like this collaborative document editor they added a month ago?
Like proton mail? The existing service that this update is adding a feature to?
I mean there’s no legal reason that would exist now that didn’t before.
My guess is that they did honor the setting, but that was because the amount of people that used it was so low vs the total number of people that used the devices. Now with smart speaker adoption rates declining, and their desire to train AI, they have to dip into the pool of people that opted not to share.