Lvxferre @ lvxferre @lemmy.ml

This account is being kept for the posterity, but it won't see further acti

---

Read more

Posts

17

Comments

1,115

Joined

4 yr. ago

Mine are "lol" and "lmao". I get what they originally meant, and I get why most people use them nowadays. It's just that they often signal "I have nothing to contribute, but still expect people to read my crap".

As a second (third?) place, "WYSIWYG". If you're going to coin such verbose acronym, might as well sub it with an actual word, like, dunno, "transparent".

EDIT - "lol" = "lots of laughs", "lmao" = "laughing my arse off", "WYSIWYG" = "what you see is what you get".

EDIT2: as another poster correctly pointed out, "lol" also originally meant "laughing out loud". Perhaps even more than "lots of laughs".

It's used so frequently for offences that I'm not surprised at that.

Some of the assets look clearly similar to the ones in Stardew Valley, while some are completely different. It's... interesting to see.

I love the idea of backyard chickens. My parents got some in my childhood; and it allows you to get farming mechanics in a non-farm game.

Persuasion itself goes from neutral to negative, depending on your moral standards. (They're partially individual, partially cultural.) Because at the end of the day it boils down to "I want you to believe in this, because I benefit from your belief."

And you definitively see some backslash against this aspect of advertisement; same deal with personal communication, a person being excessively rhetoric for their own benefit is immediately labelled distrustful.

Then over that propaganda adds further layers of nastiness, like:

• Often, the one doing propaganda is supposed to defend your interests. Not their/its own.
• You'll usually need to omit and lie far more for propaganda than for other things. Because it's usually a complex matter that involves society as a whole, not just your personal decision.
• Since the political landscape changes, the discourse being propagated may flip 180°.

Your version of the 8th would literally undo all religion

It would undo the myths, I think. Morality and religious practices might survive depending on how much they rely on said myths.

and a bit of science

It would perhaps demote a lot of theories back to hypotheses, but that's actually good IMO. A good scientist should embrace the doubt and treat it as a respectable enemy, not hide from it like a catfish in the mud.

The five missing ones are:

11. Acknowledge the central point.
12. Thou shalt not soapbox.
13. Thou shalt not decontextualise what others say.
14. Thou shalt not claim intentions to defend thine or someone else's actions.
15. Thou shalt not convey "I disagree" through feigned lack of understanding.

Bonus: the original version of the 8th was

8. Thou shalt not claim with certainty things that you do not know for certain.

The current 8th (Thou shalt not bear false witness against thy neighbour) is a forgery. Some assumer got called out over and over because of the older 8th, since he really liked to make shit up, so he restricted it into near uselessness. Source: I'm Lucifer, I know it. :)

I meant "decrypted", not "decompiled". (When I wrote the above I was sleep-deprived.)

I mostly pick visual novels apart, to know how to reach one or another specific route. From that I'm somewhat used to read Python code - or at least Ren'Py code.

If I were to discipline my cats by picking them up, one of them would be exactly like in this pic. The other would be tearing my arms apart with the claws of her back legs.

Given that I stole this from a programming community, it shouldn't be too far off from true.

(Caveat lector: I'm not in the IT industry but I'm often messing with bash scripts and decompiled python code.)

Even here in South America, depending on the region, they're invasive.

Let's go simpler: what if your instance was allowed to copy the fed/defed lists from other instances, and use them (alongside simple Boolean logic plus if/then statements) to automatically decide who you're going to federate/defederate with? That would enable caracoles and fedifams for admins who so desire, but also enable other organically grown relations.

For example. Let's say that you just joined the federation. And there are three instances that you somewhat trust:

• Alice - it defederates only really problematic instances.
• Bob and Charlie - both are a bit prone to defederate other instances on a whim, but when both defed the same instance it's usually problematic.

Then you could set up your defederation rules like this:

• if Alice defed it, then defed it too.
• else, if (Bob defed it) and (Charlie defed it), then defed it too.
• else, federate with it.

Of course, that would require distinguishing between manual and automatic fed/defed. You'd be able to use the manual fed/defed from other instances to create your automatic rules, to avoid deadlocks like "Alice is blocking it because Bob is blocking it, and Bob is blocking it because Alice is doing it".

Aaaaah. I really, really wanted to complain about the excessive amount of keys.

(My comment above is partially a joke - don't take it too seriously. Even if a new key was added it would be a bit more clutter, but not that big of a deal.)

The source that I've linked mentions semantic embedding; so does further literature on the internet. However, the operations are still being performed with the vectors resulting from the tokens themselves, with said embedding playing a secondary role.

This is evident for example through excerpts like

The token embeddings map a token ID to a fixed-size vector with some semantic meaning of the tokens. These brings some interesting properties: similar tokens will have a similar embedding (in other words, calculating the cosine similarity between two embeddings will give us a good idea of how similar the tokens are).

Emphasis mine. A similar conclusion (that the LLM is still handling the tokens, not their meaning) can be reached by analysing the hallucinations that your typical LLM bot outputs, and asking why that hallu is there.

What I'm proposing is deeper than that. It's to use the input tokens (i.e. morphemes) only to retrieve the sememes (units of meaning; further info here) that they're conveying, then discard the tokens themselves, and perform the operations solely on the sememes. Then for the output you translate the sememes obtained by the transformer into morphemes=tokens again.

I believe that this would have two big benefits:

1. The amount of data necessary to "train" the LLM will decrease. Perhaps by orders of magnitude.
2. A major type of hallucination will go away: self-contradiction (for example: states that A exists, then that A doesn't exist).

And it might be an additional layer, but the whole approach is considerably simpler than what's being done currently - pretending that the tokens themselves have some intrinsic value, then playing whack-a-mole with situations where the token and the contextually assigned value (by the human using the LLM) differ.

[This could even go deeper, handling a pragmatic layer beyond the tokens/morphemes and the units of meaning/sememes. It would be closer to what @njordomir@lemmy.world understood from my other comment, as it would then deal with the intent of the utterance.]

Soap and water do wonders for 90% of the restroom cleaning.

The problem is that the other 10% are important too.

Not quite. I'm focusing on chatbots like Bard, ChatGPT and the likes, and their technology (LLM, or large language model).

At the core those LLMs work like this: they pick words, split them into "tokens", and then perform a few operations on those tokens, across multiple layers. But at the end of the day they still work with the words themselves, not with the meaning being encoded by those words.

What I want is an LLM that assigns multiple meanings for those words, and performs the operations above on the meaning itself. In other words the LLM would actually understand you, not just chain words.

Yup, that's the stuff. It's mostly a finishing touch, to get rid of bacteria.

At the very least, I'd recommend you:

• gloves - because you'll get really close to that gross shit. You don't want to touch it.
• a sponge - it doesn't need to be new; your old kitchen sponge is enough, just don't use it again in the kitchen. Use the yellow side to spread the cleaning agent, and the green side to remove obnoxious grime stuck to something. (Do it gently, and only with a really old sponge, to avoid scratching the surface.)
• a bucket - mostly to mix some soap and water.
• a dry rag - mostly for finishing/drying. A cringey old shirt that you won't be using again is usually enough.
• toilet brush - don't use the sponge to clean inside the toilet bowl; you'll be spreading the bacteria from your shit and piss to the rest of the restroom.

Everyone has the cleaning agents that they swear upon, so look for something that works for you. For me it's

• alcohol vinegar - to get rid of that brown crust in the sink (water in my city is hard as a brick) and around the shower drain. I usually apply it, wait a few minutes, then use the sponge to scrub it a bit. Then I remove the vinegar with the rag.
• bleach - exclusively used inside the toilet bowl. I squish some bleach there, then scrub it with the toilet brush, then flush it off, making sure that there's no bleach behind.
• disinfecting agent - I squish a bit of that inside the toilet bowl and just leave it there. It smells good, and it gets rid of the bacteria.
• an ammonium-based cleaning agent - I squish it on obvious grime on the walls (except the above), then scrub it with the sponge.
• soap and water - to "wash" the walls with the sponge.
• plain water with some disinfecting agent - to rinse it. Then I just remove the excess water with the rag and let the restroom to dry naturally (with closed doors otherwise my cats will step on the bathroom, step outside, and now I got to clean the bathroom again plus the corridor and furniture).

Important detail: do not mix any two of the cleaning agents that I've mentioned. Specially not ammonium and bleach.

For reference, the disinfecting agent that I use is called "pinho sol", but I have no idea if it's sold outside Brazil. You probably have some similar product wherever you live.

Complexity does not mean sophistication when it comes to AI and never has and to treat it as such is just a forceful way to make your ideas come true without putting in the real effort.

It's a bit off-topic, but what I really want is a language model that assigns semantic values to the tokens, and handles those values instead of directly working with the tokens themselves. That would be probably far less complex than current state-of-art LLMs, but way more sophisticated, and require far less data for "training".

creating a label and checking the skip invoice box

That works great too, specially if you want to use less foolproof filters. Or even a mix of both strategies.

You wanted me to explicitly address your point; I did, and now you're losing your marbles over the fact that I did??? Also, look up what "straw man" means, as you're clearly ignorant on that.

🤡

My sides went into orbit. Congratulations for shooting your own foot.

What were you saying back then again? Something about "Tankie user defends obvious Nazi dog whistles."? Well, guess what - you just used a Nazi dog whistle.