@movq@www.uninformativ.de that would have never worked for me. I don’t have that kind of money in a bank account. 😂

We should standardise numerical notations, at least the nations using the Arabic numerals. Spain also uses the period as the thousands separator. For legibility sake, commas and periods make sense, like €3,500.25, for example. That’s three thousand five hundred Euros, and 25 Euro cents. But this should also work: €3500.25.

It makes little sense that €334.90 gets interpreted as €33490 (or €33,490). 🤯

Oof, is it any wonder some of us don’t want to just give out our info online willy-nilly.

Also that credit card ‘encryption’ will likely land that company in very hot water, no doubt far away from PCI DSS requirements.

@prologic@twtxt.net what do we make of Labor’s proposed social media minimum age ban, I.e ID verification, and the likes of Yarn? I haven’t been able to find out exactly how far the legislation goes, but some have said it’s broad enough to include any site that even has a comment section 🤔 but that could be FUD.

@movq@www.uninformativ.de The dot is the thousands separator, so I’m surprised that it did not interpret it as €334,900.00. Luckily, you caught it in time! :-)

@xuu@txt.sour.is Hahaha, nice expression. :-D

Lol. “Lighty Encrypted” https://www.pcmag.com/news/hot-topic-breach-confirmed-millions-of-credit-cards-email-addresses-exposed

Numbers are hard. I just almost accidentally sent 33k€ to someone via bank transfer, because the banking website interpreted 334.90 as 33490,00. 😬 This is germany, so it wants a comma, not a dot …

I love Al and his bass: https://www.youtube.com/watch?v=IwnDKcoVHmY

@wbknl@twtxt.net Tell us more! Who is Thomas Feiner? 🤔

Thinking about how underrated is Thomas Feiner!

You need to figure out what it means to have “free thought”, “to reason”, “have deep understanding” and be able to apply knowledge in unfamiliar environments or scenarios. You have to figure out what it means to “dream”. You have to figure out what it means to hold “ethics”, “morals” and even “beleifs”.

@slashdot@feeds.twtxt.net When are we going to resign ourselves to a position of not achieving AGI and so-called AI in the first place? Hmmm 🧐 Fundamentally I don’t think we understand how the human brain works or what it means to be a “conscious free thinking being” – I’m not convinced we’ll figure this out to be honest.

OpenAI and Others Seek New Path To Smarter AI as Current Methods Hit Limitations
AI companies like OpenAI are seeking to overcome unexpected delays and challenges in the pursuit of ever-large language models by developing training techniques that use more human-like ways for algorithms to “think.” From a report: A dozenAI scientists, researchers and investors told Reuters they believe that these … ⌘ Read more

@xuu@txt.sour.is must be extremely unlucky, as it picks a random pod in the peering list 🤣 – Oh but wait… @xuu@txt.sour.is have you updated recently? @lyse@lyse.isobeef.org did add some new feature that means you have to go to Settings -> Poderator settings and trust the peers you have there, including my pod 😅

@bender@twtxt.net True I just wanted to make sure 👍

@prologic@twtxt.net using Yarn he isn’t supposed to follow it, it should show as context, no?

They are however pretty good at auto-complete though. If you wire up Continue.dev with VSCode and a local Ollama powered codeastral model, it’s pretty decent. Or if you use the open source friendly Codeium.

@eldersnake@we.loveprivacy.club With enough data and enough computing power you can simulate anything right or create grand illusions that appear to real they’re hard to tell 😅 – But yes, at the end of the day LLM(s) today are just large probabilistic models, stochastic parrots.

@xuu@txt.sour.is You don’t follow the Slashdot feed? 🤔

@prologic@twtxt.net I never got the root for this

@wbknl@twtxt.net I have thought of getting one. I wish there were easier tools for it than direwolf

🤑😜

@bender@twtxt.net Linux and Android. I would never iOS my friend.

I like this comment on Slashdot in the above link:

LLMs don’t have an understanding of anything. They can only regurgitate derivations of what they’ve been trained on and can’t apply that to something new in the same ways that humans or even other animals can. The models are just so large that the illusion is impressive.

So true.

Generative AI Doesn’t Have a Coherent Understanding of the World, MIT Researchers Find
Long-time Slashdot reader Geoffrey.landis writes: Despite its impressive output, a recent study from MIT suggests generative AI doesn’t have a coherent understanding of the world. While the best-performing large language models have surprising capabilities that make it seem like the models are implicitly learn … ⌘ Read more

@bender@twtxt.net Ull Twt to find the references later. But my memory is quite good 🤣

I’ve been testing our new APRS digipeater, that allows us to report our geographic position via radio.

@prologic@twtxt.net sorry, tried looking for the points you mentioned, but couldn’t find them. Got a link?

@bender@twtxt.net Fair point, could be. I probably have to implement it first or create some kind of a mockup to spare me the effort of some feature that I rip out again. :-)

@xuu@txt.sour.is Yep!

@movq@www.uninformativ.de Riiiight, I now remember reading that a long time ago. :-)

@bender@twtxt.net I now read the German Wikipedia article on fog. These are some really beautiful pictures:

• https://upload.wikimedia.org/wikipedia/commons/a/a9/Nebelbank_in_der_W%C3%BCste_Namib_bei_Aus_%282018%29.jpg
  
• https://upload.wikimedia.org/wikipedia/commons/1/17/Space_Shuttle_Challenger_moving_through_fog.jpg
  
• https://upload.wikimedia.org/wikipedia/commons/9/96/Fog_Bow_%2819440790708%29.jpg
  
• https://upload.wikimedia.org/wikipedia/commons/a/ac/360_degrees_fogbow.jpg

To @anth@a.9srv.net’s points, I think this should be written as “Client recommendations” and “Serve recommendations”. Separate from the “Twtxt format” spec.

@lyse@lyse.isobeef.org

HTTP caching headers

Yes, absolutely, this should be mentioned in the spec. I’m guilty – when I first wrote my twtxt client, I forgot that If-Modified-Since is a thing. 🤦

@lyse@lyse.isobeef.org

Regarding section 4 about feed discovery: Yeah, non-HTTP transport protocols are an issue as they do not have User-Agent headers. How exactly do you envision the discovery_url to work, though?

This is from a twt of mine from January 2022:

https://www.uninformativ.de/files/twtxt/2022%2D01%2D22%2D%2Dfollow%2Dendpoint.md

(This idea gets lost all the time, so I put it into a file now. 😅)

Not sure if this is what @eapl.me@eapl.me had in mind, obviously.

@bender@twtxt.net Agreed 👍

@lyse@lyse.isobeef.org agree on the HTTP stuff. I mean we could mention that for optimization see RFC yadda yadda should be followed for caching. but not have it part of the spec proper.

@prologic@twtxt.net lies, and scams aside, there is plenty of need in the world. Yet, you can’t help everyone. Best way to do it is to donate to a worthy cause, and ignore one offs.

@bender@twtxt.net I’ve worked with this guy before. Paid him to do some freelance work. Not very good IMO. So haven’t hired him ever again. But he keeps saying hi every now and then on Signal. And then every few months or so asking stuff like this ☝️ – Last time it was money for private school fees for his child.

How am I suppose to know whether stuff like this (sound serious) is for realz or not? 😅

@prologic@twtxt.net “here, have 2 rupees.”

I need money for my mother’s heart surgery, there is a shortfall of about 2 million rupiah from a total of 40 million, can you help me with any amount?

Hmmm 🧐

😜

😜😜😜 well, now I can. This is beyond weird.

@xuu@txt.sour.is that’s on iOS? I can’t (running 18.1).

@bender@twtxt.net 😜

@bender@twtxt.net 😛

First one on the third row, for example. There are more, though.

Image

@prologic@twtxt.net how could I show you what I can’t type? 😂 Let’s see, a second.

@bender@twtxt.net such as?

I have noticed that Yarn (here) will not take some emoji. At least, will not take some under iOS.