Pup Biru

i took the phrase

You don’t need to understand why they struggle, just accept that they do.

to mean that you shouldn’t assume someone is lying. they just might have different circumstance or needs. that doesn’t invalidate their experience, just that you’re solving different problems (which may not have been well communicated, and also may not even be technical problems).

if you’re trying to solve their problems, then sure that’s a discussing… but 99% of tech conversations on the internet like this are people berating others for “not understanding” the “simple” way it’s done because it works fine for them

slight disagree: proud version is actually when you become so disillusioned with your old code that you throw it all out and start again

i moved one of those 8x2 cube storage shelves from next to my front door (where it didn’t get used) to my bedroom where i use it for “worn but not dirty” clothes

pile of clothes on the floor instantly fixed because now putting clothes “away” is the same effort as dropping them on the floor

someone elsewhere in the thread mentioned that this is the requirement, and the hooks are an “if not possible” compromise

you can’t, and shouldn’t… lemmy never claimed to be, nor has the architecture to enable it to be a private service. lemmy instances are run by arbitrary people on the internet, and some of them do run forked versions of the codebase (eg blahaj)… we have no way of verifying what’s running on the server

but interaction on lemmy doesn’t require trust. i don’t think anyone is expecting lemmy to be private

yeah, bad choice of words on my part… and i think the verification doesn’t have to be identity-based… it just has to be some limited resource (which identity is, and guarantees fairness because it’s n per identity)

it’s all compromises, and i don’t think there’s a perfect solution… what we want is the largest impact on general privacy the world over, and options that allow verifiable perfect privacy when needed - but understanding that that requires compromise in things like usability simply because it’s more complex to set up things like trust networks than to … just not

that comes down to a difference in philosophy i think… signal have detailed their reasoning for not making signals servers decentralised and self hostable, and i don’t disagree with some of them… i think everything is a trade-off, and decentralisation has scaling and usability issues

signal has done a pretty good job of creating a platform that’s much much better than alternatives in a package that’s consumable by the general public

i’m not sure that something that’s more like matrix, or xmpp, etc could do that

it might be theoretically and technically not quite as perfect, but its impact on increased privacy across the globe has been far larger because they’ve made some of those compromises

that’s reasonable. perhaps the best service is one with both options: you can somehow have a verified account that lets you msg people you haven’t connected with (perhaps they have an “allow from verified” contact option), and join groups without verification, but that you can also have unlimited anonymous accounts that are assumed spammy

that as the case may be, sending signals is still good. you don’t have to continue for very long, but a flood of support after making a moral decision will make it more likely that they, and others will make similar decisions in the future

the worst thing would be for google for example to see the fallout from this and think “well we don’t want to be them! better start building autonomous weapons”

i’d agree that for privacy alone simplex is probably better, but until it scales i’m not sure we can say that it will be able to scale. i have my doubts, simply because if you can have unlimited anonymous profiles, when it becomes a high value target then spam becomes a real problem, and then there’s only 2 major solutions that i can think of:

• raising the barrier to creating new accounts so that accounts become relatively expensive (essentially what the phone number does)
• spam filters, like email, which is a whole separate system that can be abused like it has been with email

the tracking argument against signal is pretty weak imo… signal has no ability to build a social graph because in 2018 they implemented a feature called sealed sender which is a cryptographic mechanism that allows you to send a message without disclosing to signal who you are (the receiver still knows, and rate limiting still works)

the reality of signal having your phone number is they know you (as an identity/person) use signal and that’s it

you can never validate what code a server is running, so having FOSS server code is kinda a moot point: it can’t add anything useful to the privacy conversation

the only way you can guarantee privacy is with the client code, and they have repeatable builds so you can validate the code that’s encrypting the messages, and in that case it barely even matters if their server is streaming all the data they receive to some shady other place… especially with sealed sender

one of the things i love about australian taxes: by law the displayed price must include all taxes… the price on the label/menu/sign/website/whatever is the price you pay… by law

always catches me out in the US adding tax and tip 😓 i end up spending way more than i’d intended and having a far worse experience because of it

in terms of compactness, it’s faaaaar more verbose to display time and date

i’m not necessarily disagreeing, but it’s a significant change and would likely need more than a simple text change

that solver would be tool use though… i’m talking about just the “thinking” LLMs. it’s fascinating to read the thinking block, because it breaks the problem down into basic chunks, solves the basic chunks (which it would have been in its training data, so easy), and solves them with multiple methods and then compares to check itself

what’s not how a model works? i didn’t say anything about how a specific thing works… i simply said that emergent behaviours are real things, and separately that consciousness doesn’t look like a human brain to be consciousness

given we can’t even reliably define it, let alone test for it, if true AGI ever comes along i’m sure there will be plenty of debate about if it “counts”

who knows: consciousness could just be bootstrapping a particular set of self-sustaining loops, which could happen in something that looks like the underlying technology that LLMs are built on

but as i said, i tend to think LLMs are not the path towards that (IMO mostly because language is a very leaky abstraction)

yeah i find the thinking fascinating with maths too… like LLMs are horrible at maths but so am i if i have to do it in my head… the way it breaks a problem down into tiny bits that is certainly in its training data, and then combine those bits is an impressive emergent behaviour imo given it’s just doing statistical next token

emergent behaviour does exist and just because something is not structured exactly like our own brains doesn’t mean it’s not conscious/etc, but yes i would tend to agree

also wage theft is a specific thing: failing to pay wages or benefits owed in a contract… things like not paying overtime, etc

conflating terms makes it difficult for people to educate themselves and know what their actual legal rights and options are

i’d say stability is more important than energy density

like gasoline has more than 10x the energy density than tnt and we’re perfectly fine with many kg of that on a vehicle going 100km/h

a fully fueled vehicle is the equivalent of ~600kg of TNT, but it’s very stable whilst TNT is not