brain on fire

Short answer: I don't think so

Long Answer: Maybe

The only thing I know of that you can really do is try to collect some data. The most common way is to keep a symptoms journal.

Keep track of what you eat, how much/well you sleep, how often you exercise, how often/how severely you feel fatigued, and any other potentially relevant factors you can think of.

After a few weeks you might start noticing trends around certain behaviors, or lack there of, and increased incidence of fatigue.

As for how to escape The Trap™ my best advice is try anything. Try whichever thing you feel most capable of trying in the moment.

It literally doesn't matter which thing you pick, because truthfully no matter what is causing the fatigue directly, any of those things that you can really get up and do are still going to be beneficial to you in some way.

"In the middle of a "forcing party" where friends and I are forcing one another to do the things that we've been avoiding.

So far: [bullet list] A passport has been filed for; An inbox has been zero'd; A personal website has been created; & more.

Disclaimer: I know some people are very all-or-nothing about LLMs, and that's legit and there's a lot of good reasons to be!

And also, I'd like to peer-review these tags, because I think to me personally the use of LLMs for on-the-spot, fast-and-scruffy translation is one of the cases where the benefit is big enough that the cost of an ethically-run LLM is justifiable. (Open source, permission given for training materials, fair use of renewables in a very efficient way, no personal data or text retained, etc.)

Like, someone who speaks English fluently asking the LLM to write words they could write themself in English, at enormous environmental cost, I think there's a strong argument against that. I even think there's a strong argument against spelling and grammar checking with LLMs, because they can change more than just spelling and grammar without you noticing and humans can usually understand text with spelling and grammar errors in it. (If someone needs text to be perfect in those ways in order to seem professional I would rather we do away with prejudice against people who struggle to write with perfect spelling and grammar.)

But humans understanding each other is so useful, and humans cannot learn multiple languages quickly enough that they can translate stuff online like that. Language is for humans communicating, so a Large Language Model being used to enable humans to communicate who wouldn't otherwise be able to is like, The Whole Point.

For something less transient, like a static website/sign/book that needs to speak into the brains of people whose first languages are various, it is most appropriate to hire a proper translator and pay them real money. Anything else looks really unprofessional at best! But if you're reading something and you don't understand it and you need/want to...

Like, human connection is many people's reason for being here and alive on this planet! Yeah, if we can understand each other across languages at the drop of a hat for stuff other than professionally translated texts, that's worth it! It feels very similar to single-use plastics being justifiable in a medical setting where there is no alternative, to me.

LLM is a very specific TYPE of machine learning that uses a lot more resources via wasteful computing.

Context: my partner has worked in machine learning for over 20 years (so actually knows how this all works from an insider perspective) and we also live near the site of one of the proposed mega-data centers. They are fundamentally not the same as the data centers used for traditional machine learning databases and cloud computing (I have previous posts about this but my partner explains it to me better than I can properly relay). Even if it uses renewables and even if companies weren't lying about the cooling systems being closed-loop (which as a general rule they are lying), the computing alone still generates enough of a rise in temperatures to create heat islands (coolant carries the heat away from the machines but it doesn't just vanish - it has to go somewhere and that somewhere is the surrounding environment). AI data centers, as he explains it to me, are "always-on" in a way that other, smaller data centers are not (and much of the waste from smaller data centers comes not from the end user but from ad tracking which we also would like to see eliminated quite frankly - cut that out and halt LLMs and he says we already have more than enough data centers to power the internet in perpetuity - again I trust his word on this because he's been neck deep in this field for a very long time and also has an engineering degree to go along with his experience, so he understands pieces of the hardware side that many of his colleagues miss).

Anyway the tldr of it is I'm annoyed turning AI off shuts off the translation because it suggests to me that they have eschewed tighter-monitored, more accurate auto-translation in favor of the wasteful behemoth version that's worse than what came before it (and LLMs are in fact generally worse at all useful tasks than the tech that preceded them - mistakes get self perpetuated in a way that curated closed-data systems are better able to handle).

I am in fact all or nothing about LLMs BECAUSE my partner actually knows how they work from the inside, and how they work is. bad. we are being sold a lie. Because of that this is really not the same as the single-use plastics debate. there is nothing LLMs do that cannot be done better by a less wasteful system.

How can you tell which translators are LLM and which are not?

I do not know how to tell on the surface but most places using AI for that sort of thing seem to gleefully announce it right now (I do also poke around the "about" pages when I have time to dig because sometimes you will find that info there).

Last I knew Google translate still has a non-LLM version IF you know what language you are translating from AND choose "classic" mode in the dropdown (using "detect language" automatically bounces it to AI mode as far as I'm aware).

Beyond that I would have to go hunting. My partner says Babelfish.com (that website specifically - apparently there's another one now going by that name that IS ai) used to be his go-to but we have no idea if they've starting using LLM or not. x.x It does not currently mention using AI however and its bottom-of-page copyright notice is from 2018* so it's possible, even likely, that they are still using their original software.

(*update: ok so their about page has it still at 2018 but their mainpage does bring it to 2026, but I do find it encouraging that they haven't mentioned changing tech from what they've used for decades which would be good because even in the 90s they were pretty solid.)

Here's part of their "about" page, and it looks like they use both traditional and LLM translation resources:

DeepL isn't just a translator—we're a comprehensive Language AI platform that enables organizations to communicate effectively across languages, cultures, and markets. Unmatched accuracy and performance In business communication, maintaining high standards is essential. We lead the market in accuracy by leveraging: Extensive expertise in machine translation neural networks and LLMs, tailored for business-specific translations and content creation Language AI models built and refined with years of proprietary language data A human-guided training methodology that ensures superior quality

like the first signs of a flare up from stress/over exertion used to be red face + fever + rash on my hands

but the lines of my tattoo will become raised and then a little itchy before it progresses to that point

and im discovering that, yeah, if i just listen to the wizard and rest/recuperate/stop pushing myself when it starts acting weird, i can sometimes avoid triggering the other symptoms

Probiotics as an overarching concept are not a scam. If you have recently taken antibiotics or are on other meds or anything else that can lead to gut dysbiosis then taking some probiotics could be a great idea.

Personally I only trust the kind that come in the refrigerated bottles that you get from a pharmacy (they are OTC in the US, you don't need a prescription or anything.) Probiotics are temperature sensitive so I am skeptical about how effective the fad diet probiotic sodas are and do not have any experience with those.

If you have no reason to suspect a disruption to the ratios of your gut flora then there's no reason to take probiotics and taking probiotics when you don't need them can cause gut dysbiosis. Having too much of the good critters is also bad.

We need the right amount and the right kind of critters in our guts. Commercial probiotics usually contain varying ratios of strains that are bell curved to hopefully encompass the widest consumer base. Even if you could benefit from probiotics, it's not a guarantee that you will benefit from the specific blends available.

As an autistic person with Crohn's my guts are, clinically speaking, weird as hell, and I personally never seem to fall within that bell curve and as a result had a Bad Time experimenting with probiotics. Please note that my experience does not mean that probiotics are bad.

It's just everyone is different. Lots of folks who could benefit from probiotics will benefit from the commercial blends typically available in pharmacies. But I can't stress enough that it just comes down to if there's even an issue with the ratios of your gut flora to begin with, and if the strains in the probiotic you're taking are the right ones to address that imbalance.

WOW I CAN'T BELIEVE THIS IS MY FAVORITE TELEVISION SERIES OF ALL TIME (it's not out yet)