off-topic-tech

[twin dew]

Been talked like that for past two generations of consoles constantly.

"Xbox is lagging! Xbox is therefore DEAD!"

[night girder]

Look at the views.

2.2m is a wide audiance.

Also fucking Linus.

Xbox isn't dead. Two videos below. Xbox is dying. Bro, make up your mind, click baiter.

[twin dew]

Weird to equate gamepass ultimate subscriptions to whole console ecosystem...

[night girder]

Anyway, these are the ones I think making perception worse.

[twin dew]

But clickbait wins.

[night girder]

They make content, blow it up, then move on to next topic.

[night girder]

I know right?

[twin dew]

PC GAMING IS DEAD BECAUSE MS RAISED GAMEPASS ULTIMATE PRICES!

[night girder]

Angry Joe, who I follow for ages, does it alright. Calls it game pass.

But he is also more a gimmicky guy, "angry" (so he gets angry fast) joe 😄

And he did make good points that the price change is weird. Because xbox claimed they were doing well.

5 billion in annual revenue doesn't sound bad.

[twin dew]

Revenue without costs is just fluff.
And comparison to how much they would think they would get in actual game sales as alternative.

[night girder]

Seems like microsoft took down transcript.

[soft bloom]

title likely written by some other employee or ai, hehe
funny how once enterprenour grows business and it 'starts caring about employees' customers get short end of the deal

[night girder]

Cool stuff!

You going to paint it?

[tribal kraken]

Giving it to friend whos more into that stuff. He will do finishing sanding and painting. The jaw is meant to open with your real jaw and close with rubber band. I will most likely print slightly smaller for myself and tune printer for large prints at the same time.. this was wayy too slow and fine layers to be sanded

[pure karma]

cool

[jagged snow]

https://www.npr.org/2025/10/17/nx-s1-5575254/spacex-starshield-starlink-signal

Is the link visible to y'all or did the bot block it?

[visual tree]

Great, not only are they polluting the orbit but now some of them are also causing interference....

I think more than half of active satellites in orbit are from SpaceX, we are getting closer to kessler syndrome

[jagged snow]

I must say, spaceX has been very responsible with debris and orbit management

And it's important to not that this is starshield, not starlink. It very well could've been a US millitary decision to utilize that spectrum.

[visual tree]

Probably considering those satellites are classified

[twin dew]

And they are owned and "operated" by US military, just built and launched by SpaceX

[winged valley]

I have a feeling it's a weapon to block communication with other satellites

[jagged snow]

Possibly, although that would be a weird way to do that

The smart way to do that sort of jamming would be with ground stations aimed up

[twin dew]

Not effective, as that ground station would need to be very close to the actual ground station to have any real effect.

[jagged snow]

That's a pretty blanket statement

And not universally true

You're not going to overwhelm the recievers on sattellites that way, and there are easier ways to compromise a ground station

[twin dew]

Inverse-square law is a bitch.
And if the command receive is with directional antenna, good luck otherwise.

[jagged snow]

Even directional antennae don't have magically zero gain in directions other than their foci

[twin dew]

No, but the amount of power needed to jam the actual signal goes way up if you aren't close enough in direction.
And as the distance has major effect, even without directional antennas, if you are farther from the satellite, that gets commands while over the hostile nation from their ground stations, good luck getting enough power to jam it from outside that country.

[jagged snow]

Yes, but with directional antennae forming that type of link broad spectrum jamming by just having a bunch of your sattelites transmit on that band won't do a lot either

It wouldn't be a very effective technique against much other than cubesats

[twin dew]

Still relatively more effective than trying to do it from the ground for satellites over russia or china for example.
And if the other satellite is on higher orbit, then much better likelyhood, if there is enough of the starshields in orbit, so they are relatively close to the line between the ground station and the jammed satellite.

But much more likely to just be US military trying to use band that is not that much used, fuck the frequency assignments.

[night girder]

Whatever it is. I think it's no good. Why?

1. We, the public,had to find out.
2. Country of origin of SpaceX has a track record of shaddy stuff.

.

[bronze jasper]

this is the big problem with youtube content these days. it's all algorithm fueled yellow journalism. no one does legit reporting. they just bait a negative story for a view because low brow audiences love negativity more. honest stories are boring af.

nobody wants to hear a cold take. it's gotta be hot

[winged valley]

Every major military power has a track record of shady stuff

[night girder]

IceCold, you are not the onyl one. Some content creators are noticing weird stuff in their data.
Some pointing that YT is manipulating views, making stuff more popular that isn't popular etc.

And today I watched a YT video that said the EXACT same thing as you said.

[bronze jasper]

i dont think youtube staff are intentionally manipulating views. but individual channels certainly can. you hire a bot farm to throw 10k views at your channel and the algorithm will decide it must be popular

[night girder]

There is a content creator who makes videos/documentaries about Japan. Content creator A I call him.
Then there was content creator B who made a video about how Japan is going to hell. And the content creator B shows a few streets that are dirty.
This video blew up, so every started to ask content creator A, "Hey is this true what B says, are you ok there in Japan? What's going on".

Content creator A was like wth? How did this low effort video, get so many views, and it's not even true! Because I live in that street she mentions.

In the end, they got into fight. Anway, the point I remember he made is that;

TLDR:

I can create a video, and put weeks of effort into it, make sure it's quality, fact checked, and it will get barely views.
But someone with a go-pro can make a video, that is not even reality, and that get's millions of views. Without fact checking.

[bronze jasper]

google would just demonetize a channel or something if they didn't want it shown. they wouldn't risk anti trust violations by sneakily affecting the algorithm like that

[night girder]

Yeah, let's not go into that 🤣

www.youtube.com/@JoshStrifeHayes/videos

And The Spiffing Brit is famous for breaking YT constantly. Because guess what... it's not perfect and algorithms can be manipulated.

[twin dew]

Also some time ago YouTube changed the method on how you get recommended new videos from your subscribed channels to not show all new videos by default.

You needed to change setting on per subscribed channel basis to not get algorithmicly only some, but to get all.

[night girder]

Yeah, they did some stuff back-end for sure.

The only thing I can think why this might have happened, what Josh Strife Hayes is describing, is that the views was bugged for YEARS and they implemented a fix so the data changed from august on...

So his issue is that desktop views (blue line) was always above the rest of the views.

[twin dew]

But that could easily match that change that subscribers aren't getting notified about new videos.

[night girder]

Until august, where the desktop views suddenly dip under the other mobile views.

[night girder]

No. Look at the graph. Blue, green, go up and down over the year. Always the same. More desktop views than mobile views.

And suddenly in agust, he gets more mobile views than desktop views.

And not just one of his channels.

He checked other channels, same date, same result. Asked different content creators, they have it too.

[twin dew]

Ok, yeah, those two othger types are static between, only desktop drops.

[night girder]

I would find that weird too. Can be a bug 🤷‍♂️

[bronze jasper]

i stopped bothering subscribing to channels a long time ago because youtube would never recommend me those videos anyways. i have to go to those channel pages manually if i want to keep up to date, so i just bookmark the channel pages instead

[night girder]

I was thinking yesterday; "Why do I even have a YT account?" 🤣

[twin dew]

I only log into YouTube for age restricted videos, on InPrivate window.

[night girder]

Yeah, my account is very old. I used to to collect videos I like etc.

But now if I visit a playlist from 10 years ago, half of the videos are unavailable and removed 🤣

And while I made a lost of playlist, YT recommendation is still shit lol.

[twin dew]

But basically I have never subscribed to even single youtube channel so far.

[jagged snow]

@night girder there was confirmation that a part of the drop was that some adblockers were inadvertently breaking youtube viewcounting.

That's the best explanation for why it impacted desktop so much more than other platforms

[night girder]

Haha 🤣 The revenge of the adblockers, that would be funny.

[twin dew]

Inadvertently from adblocker side, intentional from youtube side probably.

[pure karma]

Yeep Was disabling data collection and so views weren’t being counted

[tribal kraken]

Poor CPU, crunhing thru 110 million points

[cyan crescent]

Am i able to just connect my pc directly to my nas, pc Ethernet port to nas Ethernet port, and have it work that way? Just a curiosity question

[tribal kraken]

Yes, most modern network cards automatically do the needed crossover. So you dont need the special cable. But still need to configure the network manually if theres no DHCP to be reached

[storm minnow]

so im not sure if this is where to ask but i literally dont know where to go, has anyone elses satisfacorty just stopped opening after the 1.1update

[pure karma]

forget the cpu... 69GB of ram usage on the otherhand is impressive

[hidden mauve]

https://cdn.discordapp.com/attachments/1275836529880862831/1424617238002008127/IMG_4707.gif

[tribal kraken]

That also. Kinda learned the limits what can be scanned and loaded into the software before 96GB runs out and the software just closes

[tribal kraken]

I have another 96GB kit on my server, which I can borrow on case of emergency for 192 total.. Little lower RAM speed with 4 sticks neglible

[twin dew]

Configure static sized swap and it will not close, just slow down.

[dire igloo]

What else would they put as default?
There's no Firefox installed - and after you do install it, windows usually doesn't change the default on its own

Google and Apple were fined by EU because they essentially forced users to use their app stores and made it difficult to use third party ones

Microsoft doesn't do that, at least not with default Windows.
I think there is some kind of Windows which runs a reduced feature set that kinda forces you into MS Store

[twin dew]

S-mode

[dire igloo]

Ah yeah. That one

[twin dew]

But that isn't separate SKU anymore, and just setting.

[dire igloo]

Yeah, I've looked into it cuz my mum bought a used laptop that by default runs S-mode.
Seller had already disabled it himself tho

[twin dew]

Yeah, just install specific thing from MS Store to disable it.

[dire igloo]

Aaaanyways, I'm still waiting to hear from proshop.de

Ordered a 27M2N8500 on Sunday, but didn't notice it was out of stock until end of month.
430€ for 360hz QD-OLED is a steal, so I hope they don't cancel that order (they haven't yet properly confirmed it either)

[twin dew]

And for other direction, there is this setting in Settings:

[dire igloo]

btw, now that you mentioned Windows Settings:
How do I get it to stfu about logging into MS account after boot?

[twin dew]

No idea, not getting anything like that.

Settings, Personalization, Start, Show account-related notifications to Off?

[dire igloo]

Yeah, no, I don't think that's allowed.
MS will probably argue that it was a mistake - and EU would have to prove it wasn't

[soft bloom]

I forgot how bad LLMs are at geography. pretty much no use

[dire igloo]

I am deeply concerned by the implication that LLMs are considered reliable for anything but linguistic predictions

[pure karma]

They don’t force it but they also don’t let you uninstall edge in most countries

[dire igloo]

Which is fine because you're preventing the PC's internet from becoming inaccessible to the common user

Imagine you have no browser, good luck installing a browser without knowing command line shit

[pure karma]

The microsoft store that no one uses is the solution

[dire igloo]

True actually

But then you can argue that the store isn't removable

Same conundrum as downloading network drivers without working network drivers

[pure karma]

ehhh i disagree

i think the store is a function the browser is an app

i think theres a reassonable amount of things you can call part of the os but edge is not one of them

[dire igloo]

An OS can be cut down pretty hard.
It's an evaluation of bloat vs QoL

One man's QoL is another man's bloat

[limber delta]

I'm planning to purchase a passkey, may I ask of anyone here knows if the yubikey 5c nfc is a good choice?

[night girder]

If you do, buy two.

https://www.yubico.com/products/spare/

[limber delta]

yep I do plan to, but just wondering if there's better choices out there

also thankyou for the reminder

[night girder]

On the actual key, I cannot help you out since I have no experience with it. But I think a few in here have them.

@jagged snow Was it you with a yubikey?

[tribal kraken]

Oh yeah, thats what I used to do loong time ago. Is double the physical RAM good value still?

[twin dew]

Just to what is needed at max basically.
No rules really otherwise, needs to be big enough to not run out too, but extra doesn't help any.

The old rules of thumb are from times when RAM always ran out basically.

So recommendations were same as RAM to even triple the RAM.

Depending on RAM and how much drive space there was.

Basically it is just that the Windows automatic resize is too slow to react to large increases in need, so stuff gets killed.

But basically if you set too low, you still get killed.
If you set too high, you waste drive space.

[tribal kraken]

ok, but surely want it on the fastest drive

[twin dew]

Yeah, faster the drive, less the actual speed suffers.

[tribal kraken]

350GB free on C: which is Samsung 9100 pro

[jagged snow]

Yep!

What's the context?

[night girder]

👆

You got any advice on that subject?

[mental oriole]

My hate on u-boot continues... or rather this time the linux device tree spec that it uses.

[night girder]

u-boot?

[twin dew]

ARM early boot system, UEFI equivalent but much more stuff the OS developer needs to do on per platform level to support.

[night girder]

Ah, Das U-Boot.

https://en.wikipedia.org/wiki/Das_U-Boot ?

[fair swift]

never knew this was a thing the devs were trolling with that one

[jagged snow]

Sorry for the late reply, my experience has been fantastic.
I didn't spring for the 5c, just the basic security key. IMO the better option. It's about half the price, still has support for FIDO2 and NFC, can store up to 25 passkeys.
I've kept it on my keychain for going on 5 years now and had not a single issue.

[limber delta]

thankyou, sounds like a great purchase to secure my accounts!

[jagged snow]

This one specifically, if you shop on amazon: https://www.amazon.com/Security-Key-NFC-Yubico-authentication/dp/B0BVNPWPCN

[languid gulch]

🤣

[agile notch]

Ngl this the only place I can really put this but important info for any computer users!

To those unaware, Microsoft has automatically enabled a Copilot feature through Xbox Gamebar which can slow your Cpu while also sending screenshots when in-game to microsoft for data collection. Please disable this feature by going to the Xbox Game Bar (Windows button + G), selecting the Widgets Menu (the 9 dots in a square) and clicking Gaming Copilot. Once the Copilot widget has popped up, go to the bottom right and click settings (gear icon). Once in settings, click every tab and disable every feature, this will ensure every feature of copilot is disabled. While you're at it, click Delete memory and View, export or delete history. This will ensure all data collected previously is no longer being saved. Once everything has been disabled and wiped, close the Copilot widget and exit gamebar (windows button + G).

[languid gulch]

i was more thinking that Böcker apparently hired an american to do their advertising

🤣

[visual tree]

Louvre high-tech security

Their fault for having an outdated security system though

[sterile plinth]

im trying to decide whether I should go for 11" 275ppi or 12" 250ppi on a tablet i'll be using for manga/ebooks, what do yall think I should choose?

comparing it to my laptop's 150 or monitor's 90 its pretty high, but compared to my phone's 400 its pretty low so im not really sure if it would make a difference

[jagged snow]

12" is a little big for something you're using as an e-reader, I'd definitely take the smaller screen

IMO 7-10" is optimal for normal reading, 11-13 makes sense for comics

[soft bloom]

did you mean no cameras?

outdated security system would detect break in at least

[languid gulch]

i wanna know whose genius idea it was to put the french CROWN JEWELS on the 2nd floor corner of a building on a street

when i saw the Hope Diamond in DC about 20 years ago, i think i had to go thru 4 layers of metal detectors, armed security at each layer, and that was the "easy" way to it. and there were 6 armed guards around it

5 floors of Smithsonian above it, each with its own security, and literally nothing but earth below it, not even metro/tube stations

i understand that europe does tend to be more relaxed about some things than the US, but good god

if your security failure involves this at any stage, you've failed

[pure karma]

when i heard went through a window as they show the window being opened instead of broken with the ladder i immediately knew the security was a joke lmao

[limber delta]

we are at the age where the OS is almost publicly saying it's tracking our every movement now

like no shame anymore

[soft bloom]

https://www.youtube.com/watch?v=qx7hirqgfuU
this is a really nice demonstation of how deep learning works and changes depending on layers (depth, wideness) on simple example

for example of training network to predict where 'color' of a dot on a plane (actually example is a complex administrative border from IRL)
this diagram shows how rearranging same number of neurons into more layers drasticaly increases number of separate regions such odel can 'learn'
because that's quadratic versus exponential

*theoretical limits

Fudge!
it actually was 10 years...
I remember in final year of school (we have it different here) I watched his videos and paired it with a book on deep learning to not just get the math but also practice a little
A lot changed since

[visual tree]

I read somewhere cameras were aimed at the wrong direction

Corection: one important camera was pointing at the wrong direction

[visual tree]

There was probably more security than in Louvre museum when they temporarily put FIFA trophy on display here lol

[soft bloom]

who doesn't love a good tech mystery?
https://www.youtube.com/watch?v=NUAb6zHXqdI
premise is great: how come a teacher model superimpose preferences (fav animal) onto student models by trasnfering seamingly unrelated data (sequence of numbers, API endpoint code snippets)

[dire igloo]

Bruh

<@&387163995947270144> found some scam spam

[soft bloom]

https://www.youtube.com/watch?v=z64a7USuGX0
overfitting is bad at large model size is not the full picture? interesting
whole series kinda makes me want to just make some deeplearning

[soft bloom]

Why is it that anong all the stuff my pc runs, python is the one lrogram that can seemingly do nothing yet prevent sleep?..

[visual tree]

Good news: car ac compressor is spinning and there is barely any cold air coming through the vents so I suspect it's just low refrigerant

But I need to bring the car to a mechanic anyways to change filters and oil so he might as well check the compressor and residue for potential leaks

Or I think it's the ac compressor I was looking at 😅

It's right under the alternator and there is a belt probably turning it around (which is wrapped around the alternator)

[jagged snow]

Easist way to tell is by looking at the pulley for it. Usually the largest accessory pulley other than maybe power steering and it often has a very unique front face and bolt pattern because of the clutch...
But yeah, the idler portion is spinning or you'd know hehe:, clutch may not be engaged and the actual compressor may not be spinning though

[visual tree]

At least AC compressor is cheap if I have to replace it:

Whenever I hear I need to buy a new car part, it's always something expensive that costs more than €500 😅

[twin dew]

That is too cheap...
Should be in 300-400e range for refurbished units AFAIK.
And 800+e for new.

[visual tree]

That same model is €200 in a local store and compressors for my car range from €200 to €250

On amazon, it's cheaper

[twin dew]

Ok, so I'm corrected and for some cars they might be significantly cheaper for some reason.

[jagged snow]

It really depends on the model exactly, I have seen oem compressors as low as 200usd several times, especially through commercial channels

[jagged snow]

They can vary quite a lot in design and cost

[visual tree]

€800 probably if you own a Mercedes . All mercedes parts (even compressor) are extremely expensive

Not to mention yearly inspection probably costs €1000 here even if they just change the oil

But if you own a mercedes, you are most likely rich so it doesn't matter lol

[twin dew]

Meant new direct from Denso compressors.
Not car manufacturer pricing.

[visual tree]

I'm kinda surprised compressor is even running after 15 years. Don't remember we ever refilled car AC system with R134a

Although I haven't been traveling much with my car recently and it passed 157.000 km

Most cars start breaking often once they reach 200.000 km

And some break even before the warranty expires like my parent's car....

They own VW T-Roc and they replaced broken intercooler just before 5-year warranty expired for free

[twin dew]

Mainly the point was to be careful that it isn't a scam.
When it seems to be selling new Denso compressor for too low price in my mind, so might not get what it seems to be.
Might be, but might also be bad refurb, or copy etc.

Because there is large difference on how long good refurb lasts vs. bad refurb.

But it might just be that that model of compressor is much cheaper to make, than the ones I have had to look at for prices.

[visual tree]

I think it's probably that my car requires a simple AC compressor. Checked that compressor type on multiple webshops are they have almost same prices

Most expensive compressor for my car in a popular webshop is around €250

[twin dew]

What is the OEM code?

The car manufacturers own I mean.

[visual tree]

Not sure tbh, I just filtered my car (Seat Leon 1.4 TSI from 2009) on a webshop and it gave me compatible compressors

Mechanic will know which type I need to buy btw (if it's actually broken)

I know you shouldn't trust AI but

[twin dew]

But yeah, seems that that Denso DCP32045 is the right one, and is in that around 200e range currently,
Autodoc lists the "list price" as around 500e, but with discount down to near 200e.

And then other brands are cheaper.

[mental oriole]

🤣

No reality show sorry

[night girder]

You still struggling with U boats?

[mental oriole]

Considering the code is a mess. We upgraded version so ofc there is gonna be issues.

[night girder]

@twin dew does AMD X3D need xbox gaming bar?

[twin dew]

For single-CCD, no (7800X3D, 9800X3D).
For dual-CCD, preferrably yes (7900X3D, 7950X3D, 9900X3D and 9950X3D)

[night girder]

That's ... weird.

[twin dew]

Part of the core parking feature for the dual-CCDs.

[night girder]

So, if you have a 7900X3D and aren't happy with xbox gaming bar, and you nuke it, you can get issues with system?

This sounds dumb. Removing stoftware leading into hardware issues.

[twin dew]

That automatic core parking of the second CCD will not happen when running games.

[night girder]

Hardware should work independent of software, as much as possible.

[twin dew]

As it uses the Game Bar game list etc.

[night girder]

That sounds dumb. From who ever developed it. Stupid architect. Or somebody cutting corners and this was the quick/easy fix.

I don't get it. Like you literally force people to use gaming xbox bar.

Which just now got accused of data collecting (which seems false).

So if Xbox Game Bar was illegal training it's AI, users with X3D were stuck with it. Because else your CPU stops working.

Stupid.

https://www.techradar.com/computing/windows/pc-gamers-claim-windows-11s-gaming-copilot-is-capturing-gameplay-for-ai-training-by-default-but-what-its-actually-doing-is-spoiling-performance

[glossy glacier]

low effort on the side of amd but it's more flexiable in the long run

[night girder]

I don't know. They should slapped it into Adrenalin. Or find another way. Now AMD depends on Microsoft.

It's just with the recent drama about the AI training that I started to look into it. And I was thinking of removing xbox gaming bar.
But somebody warned me.

[glossy glacier]

every kind of solution needs maintenance so that it's always up to date with all games
AMD just puts the responisbility on the OS, Microsoft in this case

[night girder]

It's AMD feature X3D. It's their responsibility.

[glossy glacier]

and adrenalin wouldn't work because thats mostly a GPU software. people wont like for it to be forced either

[night girder]

It's not?

[twin dew]

It is, it is only installed in the IGP/GPU driver installer.

[night girder]

Why I said adrenalin;

1. It knows what games I play, what features etc.
2. It can overclock my CPU.

[twin dew]

So if you don't use the IGP, you don't install it.

[night girder]

Ok. Think for a second.

If AMD said; hey you ened to install Adrenalin for this CPU to work.

Or AMD says; hey you need xbopx gaming bar from Microsoft for this CPU to work ...

[glossy glacier]

thats only minor function. it is mostly a GPU software

[night girder]

I know which of the two I would pick.

[glossy glacier]

also game bar comes pre-installed. how many normal users would think of installing an additional software? how many people know enough about HW that they know why its necessary

[night girder]

And Adrenaline now already has new feature to update chipset drivers too...

[night girder]

Prebuilt builds come with bloatware out of the box. So should be fine.

People building their own systems mostly install drivers...

And if AMD tells people to install Adrenaline for CPU to work, both pre-builts and builders should be covered.

Is NVIDIA or Intel dependend on microsoft software?

And how does X3D work on Linux? You gotta install xbox gaming bar?

[twin dew]

They work without that.
Just don't automatically disable the non-X3D CCD when game is run.

[glossy glacier]

kernel does it. probably via gamemode? idk tbh

[twin dew]

Linux AFAIK doesn't have that kind of functionality at all currently.
But keeps threads of same program better on just single CCD.

But basically MS already had list of games, and functionality for Windows to change its behavior based on that list included in Game Bar.

AMD then just hooked that X3D core parking too to that existing functionality.

And easy way to mark something running as being a game.

Well, MS added hooks for AMD to do that in their driver.

[night girder]

...

Lazy. Quick. Cheap.

[twin dew]

Also, MS almost certainly had a say in how they were willing to support that feature in Windows Scheduler.

[night girder]

If Microsoft removes xbox gaming bar, which they are 100% allowed to do, the feature collapses.

[twin dew]

Because in the end it is cooperative thing between OS and the driver.
Where the OS scheduler is in control and the driver just faciliates the commands.

The HW cannot just remove some of the cores from the OS.
And if the cores are there, and the scheduler isn't part of the chain, it will not just stop using some of the cores.

Same with the Intel "Thread Director" thing, in the end that is cooperation with MS.

[twin dew]

Basically MS already had the "Game Mode" that already tied into that Game Bar game list etc.

When you use Game Mode, Windows prioritizes your gaming experience by turning things off in the background. When you’re running a game, Game Mode:

• Prevents Windows Update from performing driver installations and sending restart notifications
• Helps achieve a more stable frame rate depending on the specific game and system

So for MS, it was perfectly logical to demand that the new stuff that is very similar for the dual-CCD X3D handling to be tied to that.

[night girder]

Helps achieve a more stable frame rate depending on the specific game and system

Funny enough, I read some comments that they lose frames when they enable Xbox Game Bar.

[visual tree]

I always thought induction cooktop is the ultimate cooktop. My sister has it and my parents hate it for some reason

Guess I'll keep using the good old gas cooktop

[night girder]

I hate it.

If you have gas powered oven, you put the gas lower, flame get's lower, heat is lower.

BUT, induction doesn't work like that. If you turn down the heat, it's still fucking hot, so your food starts to boil over etc.

It's horrible to cook on in my opinion.

I don't think any professional kitchen, cooks on induction. Not famous one. Should ask Gorden Ramsey.

[visual tree]

We had gas oven in the past. We use electric oven and it's better imo. However, when it comes to cooktop, I still use gas

[jagged snow]

To be honest it sounds like crap stove
I've cooked over gas, propane, wood fire, induction, electric coil, etch

[night girder]

I want to have control over my cooking 🤷‍♂️

You all do what you want 🤣

[twin dew]

It has its plusses, but not all kitchenware work with it.
Better electric stove top, if you have the right pots and pans.
But gas has its plusses too.

[jagged snow]

I don't think any of the modern stovetop technologies are inherently better or worse from an all-around standpoint. Gas is really nice in a lot of ways but the peak output is low from most ranges and you have PM issues.

[twin dew]

Also depends if the specific unit has suitable power settings for that induction current to fit your use.
Where the larger inertia in classic electric top will not be as fiddly in a way.

Basically the induction makes the heating happen in the pot/pan bottom directly, so much faster reactions etc.

But that can take time to get used to.

[visual tree]

Do you use induction cooktop or gas one? Can't figure out what you have

[night girder]

But the electric coil doesn't cool down as quick as gas (which is just a flame).

The ceremic surface also keeps the heat.

[night girder]

I cook on gas again. Had induction 7-8 year ago. Not sure what I used back then.

[visual tree]

I use gas cooktop and electric oven

[twin dew]

Finland never really had gas stoves.
From wood to electric.

[night girder]

well then touch it with your hands. And have fun in the hospital.

[twin dew]

The pot/pan is hot, the surface can be hot.
But that internal electric coil doesn't really heat up.

AFAIK.

[visual tree]

When it comes to gas, electric or induction cooktop, electric cooktop is 100% the worst. Tested it and it's a nightmare

[twin dew]

But I can be completely wrong on that...

[night girder]

electric coil that doesn't heat up 👀

[twin dew]

Would depend on the resistance of it.
So seems they have significant resistance, but any heat at that part is pure waste.

Which anyone would want to minimize.

[night girder]

Induction heating coils 🤔 Seems safe to touch.

[twin dew]

But is that actual COIL hot?
Except by as heated by the heated part?

It gets the thing it used to HEAT very hot.
But any heating inside the coil itself is useless losses.

Which is to be minimized to get as good efficiency as possible.

[night girder]

I am not sure.

But these things are.

[night girder]

Yeah, magnetic field.

But still something has to hold the heat.

Which I guess is ceremic surface.

[twin dew]

Magnetic field is what is wanted.
Resistance losses in the coil while making that field would be unwanted losses.
And only that resistance loss will cause heating.

[night girder]

I don't think it's the ultimate cooktop, just my two cents.

When I worked in kitchens I never saw induction stoves 🤣

[visual tree]

I'll be using gas cooktop for now 👍

[night girder]

Also, gas will get cheaper

[twin dew]

Better than classic electric top in almost every way, except that need for compatible cookware.
But more expensive, and not that important when the top is on all the time.

[night girder]

The drawback of gas is, it can be dangerous? If leaks. Or you forget.

[twin dew]

But in general the residual heat for the top surface is mostly from the hot pot/pan bottom on top of it.
Not from the magnetic field generating coil.

[night girder]

Right. Well, my stuff stayed hot for ages.

You turn it full throttle, my soft boiled eggs turn into hard boiled eggs, you turn off the throttle, shit kept boiling.

[twin dew]

Also the need for the whole gas infrastructure in the city, or using movable containers you have to refill.

[night girder]

I had to take my pots and pans of the stove ... to let it cool down.

[twin dew]

But was it induction, or just normal electric?

[night girder]

Yeah, I am starting to wonder now too.

Rented the house, so it wasn't my stove.

[twin dew]

You have normal electric ones with flat top, not just induction tops.

[night girder]

It looked like this.

[visual tree]

We will probably replace gas boiler soon once we install solar panels and gas cooktop will be the only thing in the house using gas. Might as well install natural gas detector in the kitchen

[night girder]

I have extra pipeline to other house with gas. But the person who installed it, didn't use a big enought tube.

So it can't even be used for a gas boiler 🙁

Maybe I should BBQ for the rest of my life.

@twin dew , you see the red wires?

[visual tree]

Good for you. Condensing combi gas boiler are even more expensive than solar panel installation here (assuming you get government subsidies for panels)

[night girder]

And that's induction. If I can believe image.

[twin dew]

Which seems to be what the classic electric with glass top work with.
Infrared heat transfer from insane hot coils.

[night girder]

Or is that electric?

[twin dew]

All the stuff talking about induction says it doesn't get hot much, except from the heating from the pot/pan bottom.

[night girder]

Nah, you don't know the story? I think I have told it once.

They had subsidies for panels. It went totally wrong, way too popular, way too much. Goverment lost shitton of money. Next year subsidies were gone. The year after that, we all had to pay extra taxes for those subsidies.

[twin dew]

Basically the actual induction tops don't have see-through surfaces.
But those flat top glass electrics need it for the infrared transmission.

[night girder]

So now we are all paying for the fucking solar panels of our neighbours 🤣

Ah well, can be worse. Atleast it's for our planet. Right?

[visual tree]

Oh, I think I remember

[night girder]

Are we allowed to live off grid?

[visual tree]

If you have a building permit, yes

Can't even go fishing here without paying a fee first lol

[night girder]

Ok, but more like, if we all go solar power, and just have battery packs at home. Disconnected from the grid.
Then we wouldn't need to pay anyone for electricity right? It would reduce infrastructure, no more grid. It would collapse a few economies.

[visual tree]

All those cartoons or movies about people moving out in the wild are inaccurate since you can't build anything in the wild here without a permit

[twin dew]

Chinese combi-unit, left with normal electric and right with induction:

[night girder]

Is the main reason we are not allowed to do this, to keep economy stable? Or are there other very big valid reasons?

[visual tree]

I highly doubt if everyone had solar panels the economy would crash

[night girder]

I know about over generation of electricity with solar panels, which should go into batteries anyway.

Oh it would. 100%.

[pure karma]

fuel the grid with renewables that way the gouverment dosent have to pay for it

[night girder]

Like if this year, everyone swaps. World collapses.

[pure karma]

the more of it is payed by you and on your land, the less it costs them

[night girder]

The reason I think this is because I know our country buys from Germany during winter. Which buys from someone else etc etc.

[night girder]

Problem is, they will always ask money for it. They would even sell us air if they could.

"You have consumed x amount of clean air this year, here is your bill."

microchip in your lunges. 🤣

[visual tree]

Maybe the economy would crash if everyone instantly installed solar panels but if you slowly did gradual transition to solar panels, it shouldn't affect the economy much

[night girder]

Well yeah, ofc.

[visual tree]

Fossil fuel industry would collapse but there would be other industries to replace them and hire fired workers

[pure karma]

well yea thats exatly why

they get the the money for the pannels without having the pay for them in the end

[night girder]

Sorry, somebody in here will know. But was it Tesla who wanted free electricty?

oh shit, it seems it's a conspiracy https://en.wikipedia.org/wiki/Free_energy_suppression_conspiracy_theory

But doesn't seem to be Tesla.

Some people who have been claimed to be suppressed, harassed, or killed for their research are Stanley Meyer,[17] Eugene Mallove,[18] and Nikola Tesla.[19] Free energy proponents claim that Tesla developed a system (the Wardenclyffe Tower) that could generate unlimited energy for free. His system was only intended to transmit energy for free; the system's energy would still need to be generated through conventional means.[20]

[visual tree]

You mean, Nikola Tesla himself or the company?

Nikola Tesla always dreamed of giving everyone free electricity afaik

[night girder]

There was something in the back of my brain regarding Tesla and free electricity.

[visual tree]

I visited his birth house recently but it's too late to ask him if he wanted free electricity 😅 https://media.discordapp.net/attachments/457105940844380160/1296902508433903646/tesla_house.jpg?ex=68fe6815&is=68fd1695&hm=73ef0f1c10ab75130b530fff7f93d45575c6034777bab89d2958eb9678978538&=&format=webp&width=1280&height=960

[night girder]

But I cannot find anything on official wikipedia.

Only on some ... weird wiki? www.nikolateslalegend.com/wiki/tesla/nikola-teslas-dream-free-energy-for-all
I think it's a comic? Or a game?

[night girder]

I expected more something like this

a house with a massive ... 🍆

yup

[visual tree]

I loved playing with this https://media.discordapp.net/attachments/457105940844380160/1296902850101772409/tesla_coil.jpg?ex=68fe6866&is=68fd16e6&hm=be0053dd02ec10abad1b905dc4a36d53b2ee91ad4038334d333f1b9499492487&=&format=webp&width=1280&height=960

Held a fluorescent tube in my hand

[night girder]

Important detail 🤣

oh, forgot to select not but yeah 😅

Imagine him just sitting there.

Oh they demolitioned the tower? That's sad.

[visual tree]

Btw, my wristwatch battery died a day after testing tesla coil. Not sure if that's a coincidence or not

[night girder]

Haha 🤣

Tesla still messing with us from beyond the grave.

[visual tree]

I wish Edison was alive so I could punch him in the face

He was a shitty person imo

[pure karma]

terrible people make the world go round

[night girder]

and so do we 🙈

[visual tree]

Jensen says plumbers and electricians will be the new billionares. Guess I'll become an electrician if the market goes bad

The only problem is that more people will instantly become plumbers and electricians which will result in labor market being saturated with them

Like when they said everyone should become a programmer 5 years ago and now the market is saturated

[night girder]

"If we end up misspending a couple of hundred billion dollars, I think that that is going to be very unfortunate, obviously," he said. "But what I'd say is I actually think the risk is higher on the other side."

• Mark Zuckerberg on AI bubble. A month ago.

[pure karma]

the ai bubble not popping would cause way more damage yea

[night girder]

Basically, it's a risk he is willing to take. Because not taking the risk, is actually riskier?

[visual tree]

I suspect it's going to be another dotcom bubble

[night girder]

Yeah, such statements do make me worry a little bit.

[twin dew]

If it ends up being bubble, and they can deal with the losses, they continue to operate.
If they don't invest, and it ends up not being a bubble, they get overtaken and die.

[pure karma]

problem is people are actually using it at a concerning rate

[night girder]

Other people have suggested that if the bubble pops, it's a "self-contained" bubble and will not cause any worldscale harm? 😒

[twin dew]

So as long as they don't overleverage, going in is the "safe" bet.

[pure karma]

well not directly but yea the polution in the end will

[night girder]

Yeah, but wasting a few billions dollars? 100 billion as he said.

That sounds like a lot of money to me to waste on something.

While we have other global challenges.

[twin dew]

But if it doesn't kill the company, then it was just money.
Wasting money is safer than getting left behind.

As the two options.

[night girder]

For who?

Society? Or him and his company + investors?

[twin dew]

The company.

[night girder]

Why do I have a feeling, those billions dollars, are from society also.

[twin dew]

Yes, the investors will certainly feel lot of pain when the bubble bursts.

[night girder]

but don't invest any money you can't miss.

[twin dew]

But investors/stock owners != company.

[night girder]

I was wondering, if I was an investor. Would I be happy if Mark said that. Or not.

[twin dew]

But point was, if the AI doesn't end up being bubble, then if they didn't invest, the investors would lose everything too.

[night girder]

Because it means he is willing to go all in on AI, which could mean big money for me as an investor. But he also just stated he is willing to waste my money. If I would invest.

Do they*? 100 billion not invested in AI bubbled. But in other product/feature that could make profit?

If bubble burst, the money is invested in something unrelated.

Also, how far ahead is meta in the AI race? I would also take that into account. Wasting billion of dollars to catch up to competition sounds tricky.

Something still sounds off on the "could be a bubble, but let's keep pumping into it".

[twin dew]

Invest       | Lose money | Win
Don't Invest | Win        | Lose company```

[night girder]

Lose company? How? 😒

[twin dew]

By being so far behind in the race that you cannot catch up and die from loss of market relevance.

But basically that Don't Invest & AI Not Bubble is the worst case result, so Investing as long as you don't overleverage is the "safe" one.

[night girder]

I don't know. Didn't Apple work together with openAI for AI? You don't have to build one from scratch? I don't see how Meta losing AI race, would spell the end of the company itself.

[twin dew]

So that "Lose Money" doesn't end up killing the company.

[night girder]

They knew how far behind they are. So they did the smart thing I believe.

[night girder]

I think I just don't agree on the risk assessment I guess. But then I am not a multi billionaire. Or whatever they are.

[twin dew]

Yeah, but point as what the basic premise is behind statements like that.

How many large tech companies have fallen over the years to irrelevancy, because they stopped and rested on their laurels, without looking into what others were doing for "revolutionary changes".

Up to IBM etc.

[night girder]

Of the big ones? Meta, Google, Apple, Nvidia, AMD, Intel, ... they all seem to survive.

This is Meta I was talking about, they still have other products.

[twin dew]

That IBM fuckup made Intel basically.
And allowed AMD and Nvidia to come along later.

[night girder]

Like I just said; Apple is pouring billions into AI afaik. So there are other options.

[twin dew]

Intel was just one memory manufacturer for most part before IBM PC came about.
Which then others run with while IBM ignored it for most part.
And that then broke IBM mainframe business with time.

And now IBM is just shadow of what it was.

[night girder]

I understand that importance of staying relevant.

[jagged snow]

I'd say sun may be an even better example than IBM

[twin dew]

Yeah, there are lot of options, I'm just not anywhere near as familiar of what happened with Sun.

[night girder]

And I understand the viewpoint of Mark, I just can't agree with his logic. Keep pouring, because it can be worse if it's not a bubble? Let's not look for alternative ways.
Maybe innovate somewhere else? Work together?

[night girder]

Sun? That rings a bell somewhere.

https://en.wikipedia.org/wiki/Sun_Java_System

[jagged snow]

Sun Microsystems
Used to be very big

[night girder]

Yeah! I know of them. From Java 😊

[twin dew]

Which is basically why Nvidia is currently the only real winner in the AI race.

[twin dew]

Lot of other stuff too, JAVA was just side project.

[night girder]

Just saying, that's how I heard of them. Had to learn that for exams I believe.

[jagged snow]

JS was also sun origin on the programming side

[night girder]

Yeah, that's true.

[jagged snow]

They worked with netscape on that iirc
Before that also went to oracle

[twin dew]

But basically another large mainframe manufacturer that failed to adjust to x86 systems taking over the space.

As one part of the fall.

But seems to have overgone into the dot-com bubble.

[night girder]

ngl, my knowledge on the dot-com bubble isn't great.

[twin dew]

Everyone betting too early that internet would become very big.
And investing way too much into stupid shit on the internet.

And those stock market prices then burst.

Larger companies like Amazon and Cisco Systems lost large portions of their market capitalization, with Cisco losing 80% of its stock value.

[night girder]

Well, in a graph it doesn't look that bad?

[twin dew]

2.5-5.0x overvaluation bursting.
If going for that 5000 to near 1000 points.
So 80% of "value" was lost.

But basically overvaluation of stocks, and lot of overpriced buyouts etc.
And then when that burst, lot of people and companies were left with just the loan bills, when the stuff they had taken those loans for lost almost all "value".
Or just lost their stock portfolio "value", which then can cause huge amount of problems if leveraged etc.

[night girder]

Historically, the dot-com boom can be seen as similar to a number of other technology-inspired booms of the past, including railroads in the 1840s, automobiles in the 1900s, radio in the 1920s, television in the 1940s, transistor electronics in the 1950s, computer time-sharing in the 1960s, and home computers and biotechnology in the 1980s.[4]

I didn't know there were this many bubbles before.

[twin dew]

And it is getting worse, as controls on stock markets have been lost over time.
And various derivatives, futures, etc. have taken over.

[night girder]

Yeah, that's why I said, in the graph. Because I know the real world consequences are bad.

[twin dew]

But in any bubble, there are lot of people who lose a lot, but those that have money to buy into the surviving companies when they are down win.

[night girder]

About a third of the railways authorised were never built—the companies either collapsed because of poor financial planning, were bought out by larger competitors before they could build their line, or turned out to be fraudulent enterprises to channel investors' money into other businesses.[2]

[twin dew]

Seem familiar ?)

A majority of the dot-coms had ceased trading, after having burnt through their venture capital and IPO capital, often without ever making a profit.

[night girder]

Now I am not sure if I believe the argument that if the AI bubble collapses, it will be all fine.

It was one of the arguments I heard in a YT video talking about AI bubble.

[twin dew]

But basically internet stuff seemed to be the future, so lot of money was thrown on the wall to see what would stick.
And most of the stuff didn't stick, and when those tries started to fail, the bubble burst and value of even the surviving companies took temporary major falls.

As everyone was overvalued, because of the bubble.

[night girder]

Sounds familiar ...

[twin dew]

Because everyone tried to get in on the train and threw money in, causing those valuations to climb to insane heights.

Like Apple, Nvidia, Tesla etc. right now.
Where the stock price has no relation to the actual company performance.

By classic ways of comparing the stock price and company "value".

They are just high, because everyone trusts them to go even higher.

[night girder]

I have honestly no idea how big the impact of bubble burst would be.

Based on evidence from history, not great.

[twin dew]

If the AI bubble bursts, I would expect that at least Nvidia stock would crater.
But in a way that wouldn't affect the company itself at all.
But the downturn in revenue from the lost sales would cause major issues.

[jagged snow]

Sorry, what do you mean by that?

[night girder]

While Nvidia for sure would feel it, it's a bit more complex than just Nvidia.

[twin dew]

Also lot of the "private space" companies after SpaceX made it big, there are/were hundreds trying to do the same...
Most have now failed at least once, some have already failed thrice IIRC but still have found someone to buy them out and continue XD.

[night girder]

Every hopped on the AI train. Besides, maybe this one channel

[twin dew]

And now even RocketLab is touch and go depending if they can get their larger rocket out and into use fast enough or not.

When SpaceX ride shares have taken much of RocketLab Electrons market.

[night girder]

SpaceX is getting so dominant. They also manufacture starlink ...

on the bright side, EU is a go for satellite network I believe.

[twin dew]

Which is large part why they got so large in that smaller launch space too.
Because Starlinks are currently payloads of most of SpaceX launches.

[jagged snow]

Starlink brings in more revenue and profit for them than launch at this point

[twin dew]

If they didn't launch Starlinks, Falcon 9 launches would happen way less often, and the costs per launch would be much much higher (not price).

[jagged snow]

They're an internet company with a side business of medium and heavy-lift launch

[glossy glacier]

It does in jacascript /s

[night girder]

Are there any rules/laws for who can put satelites where?

Like, do countries get assigned orbits?

[jagged snow]

There are a lot of international agreements

But not assigned orbits for countries

[night girder]

Because now I am thinking, great that EU wants to launch more satellites into space, but what if they collide or something.

[twin dew]

Not for below GEO, GEO has strict slots.

[night girder]

Like, it's only getting more busy up there every year 🤷‍♂️

[twin dew]

Which is why the various constellation companies don't even try to use same shell altitudes.

For US, FTC had to approve the shell heights for Starlink etc.

And make sure they wouldn't be danger to others, and to keep SpaceX in check for that.

[night girder]

Ok, but internationally?

[twin dew]

Cooperation between the regulators basically.
And ITU handling frequency spectrum for some part.

But basically SpaceX has now "dibs" on 340–360 km, 525–535 km and somewhat on 604–614 km circular orbits.

[night girder]

Apparently collision avoidance is done by the operator.

When two satellite operators are notified of a potential collision, one or both operators may decide to maneuver their satellite, eg. ESA & SpaceX in 2019.[11]

[twin dew]

Yes. With international coordination for that.
Traditionally there were so few satellites, that satellite-to-satellite wasn't really issue for most part.
But basically no-one will really want to go into same height as one of the mega-constellations now.
But as they still take just small part of the possible "bands", that isn't issue yet.

[night girder]

Apparently it did happen in 2019.

[twin dew]

Amazon Kuiper in turn is going for about 590 km , 610 km and 630 km shells.

[twin dew]

Yes, that was orbit raising SpaceX satellite.

[night girder]

If I was ESA operator, I would have gone after that SpaceX satellite 🤣

[twin dew]

IIRC.

[twin dew]

ESA didn't trust SpaceX to execute their avoidance maneuver correctly.

For whatever reason.

[night girder]

SpaceX, in a statement Sept. 3, said it was aware of a potential conjunction Aug. 28 and communicated with ESA. At that time, though, the threat of a potential collision was only about 1 in 50,000, below the threshold where a maneuver was warranted. When refined data from the U.S. Air Force increased the probability to within 1 in 1,000, “a bug in our on-call paging system prevented the Starlink operator from seeing the follow on correspondence on this probability increase,” a company spokesperson told SpaceNews.

[twin dew]

Ok, I then misremembered.

[night girder]

“SpaceX is still investigating the issue and will implement corrective actions,” the spokesperson said of the glitch. “However, had the Starlink operator seen the correspondence, we would have coordinated with ESA to determine best approach with their continuing with their maneuver or our performing a maneuver.”

So, either cover-up/damage control. Or glitch.

Or maybe the Air Force fucked up. 🙈

[twin dew]

And because of that, SpaceX started to publish very accurate orbital data for all their satellites publicly.

[night girder]

Fair of them.

i forgot when he was in charge and when he wasn't.

[twin dew]

He has always been at charge at SpaceX, difference is just how hands-on he is with whatever part of the company at what time.

[night girder]

that's what I mean with in charge.

Not the boss on paper 😉

The actual person(s) making the descisions and putting them in practice.

I mean a few months ago he was in the goverment no? 🤣

and then he has X.

[twin dew]

Starlinks are on too low orbits, outside of that planned about 500 satellites at around 610km, for that.

[night girder]

Can't be hands on and be everywhere at once.

[twin dew]

Amazon Kuiper is in the danger zone for that height.

Going for that higher height to need less satellites and for them to last longer.

[night girder]

Also, I don't think he was angry. Nothing happened. They made evasive maneuver. Made some public statements and that was it.

[twin dew]

From Wikipedia.
And lighter a thing is compared to its surface area, faster it comes down.

[night girder]

How many corrective maneuvers can these satellites take?

[twin dew]

Just uses tiny bits of Delta-V, which is needed for altitude raising anyways.
So mostly just raise altitude tiny bit now, and not later.

[night girder]

Ah, so nothing really lost. Efficiency wise.

[twin dew]

But point was, that stuff at 500km orbit decay still in just years, while stuff at 600km orbit decay in decades.
Which is gigantic difference.
And that lowest Starlink shell at around 350km is in months.

China too, and India has local system, Japan is planning for their own too.

Local as in few satellites in weird orbits so that the satellites spend most of the time over right area.

https://en.wikipedia.org/wiki/Indian_Regional_Navigation_Satellite_System

[night girder]

I was wondering why.

But I read the wiki and I know why now 🤣

[twin dew]

Higher the orbit, less satellites you need for same coverage, but higher the power needs and more there is latency.

[night girder]

The system was developed partly because access to foreign government-controlled global navigation satellite systems is not guaranteed in hostile situations, as happened to the Indian military in 1999 when the [INSERT A COUNTRY HERE] denied an Indian request for Global Positioning System (GPS) data for the Kargil region, which would have provided vital information.[22] The Indian government approved the project in May 2006.

[twin dew]

Ah, that.

[night girder]

I was genuinly wondering what the reasoning could be. I know about being independent.
I had hoped maybe they had another reason I didn't see. Efficiency or something else.

[twin dew]

Yeah, same with EU going with Galileo, not completely trusting US with GPS.

[night girder]

No. We had proof. They cut it off at one point 🤣

That's why we are getting our own satellites, hopefully.

[twin dew]

Civilian or military level?
For civilian side it IIRC was more about not trusting US to not have a way to degrade civilian signal anymore.
And for military side not being willing to bet on US always giving military signal level access to other NATO countries.
Not about them ever cutting anyone off.
Basically "Better to have it and not need it, than to not have it and need it".

[night girder]

We need our own starlink network. Simple as that. The past has the proof.

[twin dew]

GPS formerly included a feature called Selective Availability (SA) that added intentional, time varying errors of up to 100 meters (328 ft) to the publicly available navigation signals. This was intended to deny an enemy the use of civilian GPS receivers for precision weapon guidance.
Which officially hasn't been supported in any Block IIF or newer GPS satellites, which have launched since 2009.
And has been off since 2000.

[night girder]

My phone is pretty acurate I think.

Or do you mean the feature is turned off?

[twin dew]

Yes, that it was turned off in 2000.

[night girder]

maybe we turn it on if we are at war 🤷‍♂️

[twin dew]

The civilian GPS signal alone is officially accurate to about +-10m, in actuality with current simple implementations to about +-1m in open.
But that SA added +-100m degradation to the signal timing at source.

While it was on.

[night girder]

Well, satellites at this moment are accurate enough for human targets. We know that for sure.

[twin dew]

Basically making the civilian signal 10x worse than it was technically.
GPS military signal is 10x more accurate than the civilian signal on technical level.

[night girder]

They track phones to hit military targets.

Pretty darn accurate.

[twin dew]

But needs encryption key to decrypt.

[night girder]

They use satelite images anyway. Even if it's innacurate. They use the image to fine tune the target.

[twin dew]

And as the military and civilian signals were on different frequencies, you could use both to get even better accuracy than either alone, if you had the decryption key.
That +-1m civilian accuracy from +-10m civilian signal is because of using unencrypted military signal in the mix now.

[night girder]

Locate by phone signal to get radius, scout that radius either with images from satelite or drones.

[twin dew]

And then also using the other constellations now.

Basically someone realized you can use the military signal, without understanding the contents, to compensate for some of the errors in civilian signal propagation.

Caused by atmospheric effects, which are frequency dependant to some extent.

[night girder]

Sounds cool, a bit outdated 😄

wasn't it @visual tree whos garage door opened because it was on a military frequence or something? 🤣

[twin dew]

Also now ground stations giving correction factors over internet.
Basically very accurate GPS receiver on the ground with exact known location.
Reporting the difference between the GPS signal it receives and the actual location as offset for other devices to use.

[night girder]

Also, can't you use multiple satelites to get radius and calculate on those multiple radius where they overlap to narrow target?

a bit how they triangulate cellphones?

[twin dew]

Lot of ways current receivers go better than the original "theoretical".

[visual tree]

Yeah, I had a garage door motor so reliable nobody expected it would last 15 years. Bought it during the time when electronics used a frequency which was later used by military 😅

Now, you can't sell electronics designed to use military frequency

[twin dew]

Yeah, but back then that was civilian allocation.
That just at some point was cancelled for civilian use, and then reassigned to military after transition period.

[visual tree]

This 👆

I remember when I called garage door company to check why the door opened by itself during the night and they got shocked when they realized I still owned a working american garage door motor (the brand was Genie)

[night girder]

Lucky you were never at war and they considered it a military target 🤣

[twin dew]

Ah, US one, might have used US frequencies then.
As US and EU have different allocations for stuff like that.

[night girder]

which seems to be public information ...

[pure karma]

considering a garage door motor as a militairy target would be insane

[visual tree]

It's actually tragic they got surprised it was still working after 15 years. Shows you companies know modern electronics don't last as long as they used to in the past

[night girder]

if it's on a military frequency?

[twin dew]

Only the opener transmits, the garage door side is just receiver.

And only when you press a button.

[night girder]

still, if they catch it 🤷‍♂️

[visual tree]

Don't have to worry about it anymore. It's probably somewhere in the recycling plant

[night girder]

or somebody else 🤣

[visual tree]

It's broken so you can only use it for parts

My case is not the only one though. When Pope John Paul II visited Croatia in 2003, there were numerous reports of people's garage doors opening by themselves

They also probably had genie motor and the military frequency interfered with it

[night girder]

It sounds like something out of a comedy. 🤣

[visual tree]

Yeah lol. Not going to happen today since almost all old electronics that use current military frequency are probably broken

[night girder]

Yeah, and it's public what frequencies are a no-go.

This one even says on what frequency nuclear detection is done.

1215-1390

[twin dew]

Yeah, seems that in US, 40MHz is very legacy, but 310, 315 and 390 MHz are all common use for garage openers.
But all three later ones at least in EU are "Defence systems" and licensed personal radios.
40MHz is more free, but still includes also military stuff, but with things that might confuse the old opener to open when not intended.

While in EU, 433MHz ISM band is where stuff like that is supposed to live.

[visual tree]

tablice.hakom.hr:8080/vis?lang=en (Radio Frequency Allocation Table )

Shows you exactly what frequency is used for

[twin dew]

The actual original source: https://efis.cept.org/sitecontent.jsp?sitecontent=ecatable

[visual tree]

Today, 328.6 - 335.4 MHz range is used for Aeronautical radionavigation and 335.4 to 390 range is used for mobile

[twin dew]

That is partial take:

Just takes.
that "Mobile" just means that the things using that band aren't fixed in place, but can move about.

Doesn't mean civilian mobile phones etc.

[visual tree]

Oh, my bad

[twin dew]

PPDR = Public protection and disaster relief

[visual tree]

Learned about 433 MHz when I bought a sensor for my alarm system only to find out there are two versions: 433 MHz and 868 MHz 😅

[twin dew]

Yeah, that 868 MHz seems to be another common use ISM band in EU.

[mental oriole]

That view is not very intuitive...
I thought...
...Wait aeronautics use 0-90khz, the rest is used by mobile wtf is mobile using all that range for.

[twin dew]

Like I said, not "mobile" as in "mobile phones", but "mobile transmitters"

That site gave very incomplete information.

Just the very high level use type allocation.

[mental oriole]

Guess I'll have to make my point even more clear.

[twin dew]

And do note that this pic is for that tiny bit marked in red.

And mine are snaps from 284 page document for 8.3khz to 3THz range allocation in EU.

[mental oriole]

This is how I read the ranges.

The black lines added for clarification

🙃

[twin dew]

That wasn't it.

[night girder]

whats happening.

why is Baldur drawing pyramids.

[mental oriole]

I'm complaining that the view is very confusing if you just glance over at it.

[visual tree]

Yeah, wish they made a better design

[mental oriole]

[twin dew]

Yes, bad site.

But I'm off to bed, should have been there 2+ hours ago.

[night girder]

Good night 👋

[visual tree]

Oh, it's daylight saving time again....

[night girder]

It is?

[mental oriole]

tomorrow ye

[night girder]

Ah, cmon.

[visual tree]

I mean, we have to move clock by 1 hour

[night girder]

I don't want to time travel.

[mental oriole]

Thanks for reminding me

[pure karma]

but its dark.... 🙁

[night girder]

it's always dark somewhere 🤷‍♂️

[mental oriole]

My heart.

oh wait

[night girder]

🤣

[pure karma]

yea well 8am isent somewhere 😂

[visual tree]

EU opened a discussion regarding abolishing DST for the 100th time and I have a feeling nothing is going to be agreed again like the last time

[mental oriole]

Like always.

[night girder]

Wait, didn't they agree, to allow it, if we could agree what timezone everyone switches too?

and that's the hard part. Everybody needs to be onboard to pick timezone and take into account neighbouring countries etc.

[visual tree]

That's the biggest problem, everyone is arguing because each country wants either winter or summer time

[night girder]

So EU decided, the let the members sort it out 🤣

[night girder]

Yes, that's the last I heard of it.

We should just say fuck it, and pick one. A tiny bit of chaos.

I guess operating systems etc must all get updated too.

[visual tree]

Someone made a proposal

[night girder]

I don't think someone is going to agree ... looks all the way too the right.

[visual tree]

Maybe I shouldn't have posted that proposal. Now, everyone from europe is going to be angry on discord 🤣

[night girder]

Eh, I just want to get rid of the time shift every year tbh.

[visual tree]

Same. I'm tired of this DST shit

[night girder]

It's overly complex, I think the science behind it is outdated, ...

[visual tree]

It worked in the past but not anymore, it's useless now

[night girder]

Kinda funny how we went from this

To this

It started simple ... and then it wasn't 🤣

[mental oriole]

I'm gonna create Kit time

it's always 01.00

[night girder]

can I join your timezone? :3

I'll be a kit citizen.

[mental oriole]

[visual tree]

We could also freeze time if something goes wrong

[mental oriole]

Kitizen

[night girder]

I am not paying taxes. Let's get that out of the way.

I blame asablic.

[dire igloo]

Wait until you see Australia

[dire igloo]

My coding professor during 5th semester said that he'd give any student the highest possible grade if we managed to have our own time zone recognized by some official body

[mental oriole]

[fallen cradle]

Tech 👍 (£25 from a car boot)

[visual tree]

Don't remind me of selling working NES which my dad bought in Japan 😭

After 20 years it was still working! (maybe it was much older)

[soft bloom]

and YT strikes again!
https://www.youtube.com/watch?v=2HY-nREvVu4
might be some sort of A/B test or any other flavour of deploying feature for small number of audience instead of for everyone at once

are there alternatives to YT?
wall, for entire life, except for some gaming oriented video hostings, and subscription based stuff like Nebula, Curiosity stream...
I could only remember Vimeo and Ramble haha.
YT's monopoly is so large that most people don't even try to compete or look for alternatives

[twin dew]

There aren't direct competitors.
All others require either payments from the uploader, or the watcher.
Utreon/Playeur is now dead.

And even that topped at 720p for free viewers.

Everyone who was free to watch before that required payments from the uploader instead.

And all the "competitors" now in operation require monthly payments from the viewer.

[dire igloo]

You could also use other social media platforms that aren't meant for video hosting.

Instagram or Twitter could theoretically work, but their respective frameworks just won't make it as intuitive for watching videos as YouTube

Hell, even Facebook could probably work

[tribal kraken]

Vimeo?

[manic cipher]

Try Rumble.

[twin dew]

Needs subscription from the uploader.

[glossy glacier]

I'm thinking about a nebula subscription

[inland onyx]

why? wouldn't they all just continue using Unix time anyways

[jagged snow]

Formatting for dates would still need updated. In some cases it's fetched but in many it's programmed statically.

[soft bloom]

The only advantage of shifting time, is that people can keep alarms set to same 'time' and go to work on same 'time' without shifting their actual wake up/sleep actual time relative to sun
But who the fuck benefits from that?
It6not hard to make alarms follow sunrise sunset time based on preset geolocation.
Convenience of knowing that stores open at 9 am is diminishingly small compared to having better health necause you don't have to change sleep schedule 2 times a year potentially causing all sorts of health problems.

It was a temporary fix that somewhat worked and stayed for far too long

[soft bloom]

in the context of sharing unrecognasible to humans strings (texts, emojis etc):
https://www.youtube.com/watch?v=wm642z2048c
so they share 'glyphs'... which means, that through the mechanism described in Welch Labs video, they are doing the same thing - aligning AI through seamingly nonsence data

[twin dew]

There are updates to the timezone to country mappings usually at least once per year, whenever some country in the world changes things.

So that part is non-issue.

Changes exactly when they change between summer and winter times etc.

Debian version of that timezone package for linux and its changelog for stable version:
https://metadata.ftp-master.debian.org/changelogs//main/t/tzdata/tzdata_2025b-0+deb12u2_changelog

So Linux side can even handle individual regions inside countries doing their own thing:

tzdata (2025b-0+deb12u1) bookworm; urgency=medium

• New upstream version 2025b:
  • New America/Coyhaique zone for Aysén Region in Chile, which moves from -04/-03 to -03. It will not change its clocks on 2025-04-05.
  • Improve historical data for Iran
• Add America/Coyhaique to debconf templates

https://en.wikipedia.org/wiki/Tz_database

So even if each EU country did its own thing, without any care of the neighbors, operating systems would just update that DB to newest version and go on.

And as each EU country already has its own TZ entry, no user side config necessary to change anything needed either.

[twin dew]

OCCT gets test mode where it intentionally tries to create coil whine "music", to make it easier to determine the location.
https://www.tomshardware.com/pc-components/gpus/occt-version-15-adds-coil-whine-detection-that-doesnt-require-a-microphone-popular-stress-tester-gets-genius-new-feature-to-silence-your-pc

[soft bloom]

haha, nice

[visual tree]

This? 😅

Guess you could say Australia invented time travel

[dire igloo]

Yeah. Their DST/non-DST time zone shenanigans are wild

[dire igloo]

I read this as "Arizone does not deserve daylight savings time" lol

[languid gulch]

oh we don't

half the year we're with LA, the other half with denver

phoenix never changes clocks

oh brain fart, misread yours 🤣

4am brain

[dire igloo]

XD

[visual tree]

Good morning or good evening then (depending if you woke up or went to bed )

[languid gulch]

oh this is my sign to sleep 🤣

[dire igloo]

And I just realized I misspelled Arizona for some reason and my autocorrect didn't catch it

[pure karma]

Morning

[cyan crescent]

Since i dont have a gpu in my nas, could i use the gpu in my pc for steaming somehow or no?

[dire igloo]

Interesting question, kinda wanna see where it goes

[twin dew]

Only relevant if you would want to do transcoding on the NAS.

[dire igloo]

If I have a video file on my Nas and I open it on my PC without transferring it first, how does it transmit?

[twin dew]

Changing what format a video file is in, either after saving to NAS, or while streaming from the NAS to device that doesn't support the format the file actually is in.

[dire igloo]

nvm, just gets loaded into temp storage locally

[twin dew]

Depends on how you open it.
Via file sharing protocol, or via DLNA or like streaming protocol.

That transcoding is thing for streaming protocols where the program acting as the server on the NAS does transcoding on the fly.

[cyan crescent]

I might try to stream from my nas to another state

Pc will be on the same local network tho

[twin dew]

And not usable, if the PC isn't the computer running that streaming server program, and accessing the files via file sharing protocol.

At realtime.
You can always transcode on the PC, and save those transcoded files on the NAS for the streaming use.

[cyan crescent]

I have a bunch of bluray files i ripped. Dont think those will work well without transcoding

[twin dew]

Depends on network speed.
But likely not.

[cyan crescent]

I only have 100mbps up from the nas to the internet

[twin dew]

So what CPU and OS do you have on the NAS?

[cyan crescent]

Trunas with a 7600x

[twin dew]

And that has IGP in it, that has AMD decoders and encoders.

Same as any AMD GPU would have basically.

[cyan crescent]

How good are they?

[twin dew]

For RDNA2 gen, or between RDNA2 and RDNA3, as sometimes the IGPs have half-gen updates to that video block.

[glossy glacier]

7000 series should be RDNA2

2CUs

[twin dew]

Yes, the cores themselves are RDNA2, but didn't check if the video block got update over the dGPU dies.

That can be different on same main arch.

Yes, little updated.

NAVI 21 and 22 use Video Core Next 3.0.0
NAVI 23 uses 3.0.16
NAVI 24 uses 3.0.33
Raphael IGPs use 3.1.2

HW versions for the fixed function video block that handles encode, decode etc.

RDNA 3 and RDNA4 dGPUs both use Video Core Next 4.0.x.

Basically that 3.0.x vs 3.1.x means there were some actual functionality upgrades, but not enough to count as full generational change.

[glossy glacier]

probably just some minor changes to resolution per codec

or optimizations

[twin dew]

But the important thing is that H264 encode B-frame support already came in VCN 3.0.

[soft bloom]

That vertical difference is... Unusual
Like really why
Unless they are so far south that curvature plays role

[neat pine]

Urgh, vibe coded nonsense.
My mum is a million years old with failing eyes and is trying to sign a contract for a new bathroom build
The builders sent her a contract by email and she can't read it on her screen, she wants to print it so she can read it properly before signing
But it's some weird-ass bespoke interactive format that can't be printed and screen readers just give up with it
She wants me to help from 30 miles away
Not looking for help or support here, just ranting, I'm gonna have to go over there after work and figure this out in person

[dire igloo]

As much as I hate accessibility regulations at work, this is why it has to be a thing

I wonder how enforceable that contract would be had she agreed to it without being able to read its terms

[soft bloom]

Accessability is a thing that can be annying only until you are the one needing it.
Aside from half-assed implementations that help noone.

How many people here have some sort of adhd and just created a tiny bubble around of them tht is accessible to them but is completely unfathomable for most other people?
Some could argue thta whole tech, or gaming is like that.

Good UI exists only because someone implemented lessons from what works for most. And on top of that we often need more stuff to help cover remaining people.

Any drone experts here?
I wonder if there are any ways to improve visibility in harsh conditions when you are trying, let's say... Home in on unmanned aircraft.
Like, infrared probaly will dissapate too fast.

But I guess one can alwyas use that trick with nicrophones and dopler
Except idk how well it will work from the rear

Just that after what happened coyple days ago been entertaining idea of how to rpevent that shit

Well,.not just from that event. But it reingnited the process

[dire igloo]

Guess my fpv knowledge ain't too helpful there tbh

[dire igloo]

Intuitive UX is one part of accessibility that ironically is often overlooked making "certified accessible" software less accessible to people without physical disabilities

There are certifications for accessibility that look at stuff like screen reader compatibility, being navigatable with only a keyboard, having proper readability (contrast, UI size, font choice), UI explanations, etc

BITV 2.0 is one that I've come into contact with a couple times

Issue is: accessibility for people with mental disabilities is often overlooked.
And because UX design is so easy to save money on, you get a lot of "low barrier" software that you have to have a hundred page manual with you to understand

[soft bloom]

I mean, fpv is cool, but are you going to see anything in the middle of the night?

[dire igloo]

Fuck no, not without lights

[soft bloom]

I wonder what range on 20W LED would be

or rather if dark paint would reflect back anything noticable

on the other hand, running some microphones array and transmitting that data with video feed to compute sound sources on receiving end... that could enhance experience

so that video is more for visual confirmation

and then maybe some net with parachute to avoid more boom boom stuff

I saw some fpv can reach high speeds but wonder if they can still compete with jet engine drones 🙂

[dire igloo]

Always.
But the actual question is whether it's noticeable: and that depends on a lot of factors.
strength of the light source, distance to the object, absorbtion coefficient of the paint, focus length of the beam, absorption coefficient of the atmosphere, sensitivity of the detector, atmospheric scattering, etc

[dire igloo]

The point of fpv is short range, high speed, high maneuverability.
It's a hobby, not intended and not suited for warfare.

You can adapt hobbyist drones for warfare, but they quickly become unfit for FPV racing

I doubt jet engine drones could navigate an fpv racing parcour.
idk if that'd still be the case if you scaled up the parcour to adjust for size difference, tho I'd say it would

[thin trout]

she is herself infamous for pseudoscience

[soft bloom]

like what?

[thin trout]

Lots of stuff e.g. https://old.reddit.com/r/AskPhysics/comments/1isje08/what_is_the_general_consensus_of_physicists_on/?sort=top and pseudoscientific hate of minorities.

[dire igloo]

I can give an example of pedaling conspiracy theories:
Particle physics is useless, it only exists for physicists to rake in funding money

I prefer the Miniminuteman approach to science:
If science feels like an ivory tower, the solution is not to remove scientists or to defund their work but to establish good science communication

imo, the majority of "this research is useless" criticism comes from a lack of understanding and not because it's actually useless

And even if it was useless, someone else might pick up that work and accomplish something very useful.
That's why the Ignobel Prize is so important imo

[soft bloom]

and the only 3 matches for "pseudo" are not related to her? doesn't prove the point for "pseudo scientist"

[cyan crescent]

Can i test hard drive speed on trunas?

[soft bloom]

that's missing the point completely.
from what I heard her say, she wasn't picking at research that doesn't seem necessary at the moment, but particularly criticized kind of research that is not advancing our understanding of... anything.
it was in one of her most recent videos on that topic...
let me rephrase: there's a bit difference between writing 1000 papers on "what is the result of multiplying 1232134217834 by 963096324059?" and writing 1 paper on "what if numbers had imaginary component?"
you can make mane many 'reasearch' of the first kind, but it's the second kind that will give more push to the science

[soft bloom]

that can be true, but again, that wasn't her (how I understood it) point

[dire igloo]

Except her point was "we're investing billions into more and more particle physics research tech, new colliders, new labs, etc. But nothing we do there matters, all they say is 'this helps us understand the universe' but they're just running a big con to get research funds"

It's the "mathematics is fake" meme but from someone who can act smart

[dire igloo]

Because instead of calling her a pseudoscientist verbatim, criticism is best brought up by mentioning specific issues you have with someone.

You don't need to refer to someone a pseudoscientist, you can just call them out on the things they do that are considered pseudoscience.
And when you do that, no Ctrl+F "pseudo" in the world will match your comment

[thin trout]

conveniently ignoring the pseudoscientific hate towards minorities too

[soft bloom]

mentioning minorities doesn't ring any bell for me, so I have no idea

[soft bloom]

ok but then what I read so far doesn't summarise to pseudo science

[soft bloom]

no the"minorities" but her involvement, and especially hate

I mean, if she was wrong, she's no more "pseudo science" then journalists falsely calling out fake news

anyway, I remember linking her videos to people in the field, and they generally agreed with her criticism.
There's some alignment (and therefore bias) in my views on science before heard her, but she did put some nuance into the question.
I think it's right to put some thought into where budget is spent, and how.
tax payers have equal right to ask for results from politicians, scientics, and any other government/tax funded projects. side eye at military
I am all hands on increasing budget for science, but it will only make things worse if that budget won't make for any increase in yield: breakthroughs and consumer technology.

[dire igloo]

Criticism that I've seen the reddit discussion mention:

• baryon asymmetry is nothing important and needs no explanation
• unification of gravity and quantum mechanics isn't possible and should therefore be defunded (no proof)
• pedaling LIGO scepticism by spreading misinformation (debunked article/paper)
• misleading claims about the usefulness of particle accelerators/colliders
• promoting transphobic viewpoints

And just a general criticism of appealing to a right wing crowd through the same means as other grifters

According to one comment, there's a multi hour long video series of Professor Dave Explains criticising her content going back and forth with her responses
Probably a good watch, I've seen bits of his other content

[inland onyx]

I love his back and forths with the flerfers

[twin dew]

So basically another nut who selects position first, and then finds stuff to match that, ignoring anything countering.

[dire igloo]

Two things to note about that:

1. She's not providing alternative programs that should be funded instead, so she reinforces the right wing grifter motion of defending science in general
2. Some fields like particle physics and astrophysics tend to slow down in discovery speed. Tieing funds to new discoveries promotes publication bias or kills those fields

[dire igloo]

Many commenters agree that she started off as a solid science communicator with valid criticism about prioritizing funds and ivory tower academia but that she then drifted off into fields she had little expertise in and later deranged into sensationalist grifting

[soft bloom]

so coming back to...
that's a far stretch to compare ai awaking that uses physics/math terminology for roleplaying
to same tier of 'pseudo science' as what her criticism and drama claims

[soft bloom]

1. Yes she doesn't... Except somewhat suggesting that research should be done differently? I can't remember a concise quote but she had some dieas voiced of what could work better.
2. Yes, but again, there's difference between say documenting all 10000 backyard toad camouflages and making expedition into deep sea where nobody was before.

Like, come on, we all know there plenty of redundant science projects that just don't do any good. My personal experience from university is maybe somewhat outdated, but considering that it already lagged behind in decades, I doubt much changed after covid.

Problems should be pointed out. And I don6 are if that somewhat aligns with bad actors. (Mostly because not involved in us politics as much)
Some shaking is ofcourse not a thing of stability but hopefully it will be the push for science as a whole to focus of what can proof it's worth. (But we all know that scientists worrying about food is mostly not an efficiency boost)

Who doesn't know about struggles of making your research/project look relevant?
I got US money in the past. I know what buroucratic hell is applying for grant money. I know how much headache is in justifying what you do is owrth it.
And honestly, I don't think that without specific language and "tricks" there's any chance to get approval.
(That was before covid though)
Way too much time was spent on dancing around papers. Time that could be spent on making project better.

Anyway, I mentioned what she commented on the news,
Got unexpected recation and claim attached to her, we went deeper,
I talked about what she said, I defended some atatements.
But it's not that I necessarily defend her as person, am much more interested in topics she touched.
As always, no matter who says that sun rises, it still exists.

[dire igloo]

Redundancy in science is important, it creates confidence.
We not only need to advance our knowledge but also verify it.

And what I meant with number 2 is more along the lines of "at some point, cosmic events happen so rarely and particles interact so little, that you can't do anything but scale up detection".

And yeah, I recognize your repetition of something that sounds like something she said to delegitimize fields of research.
But in the end, "what happens if we add these two things together" is still a crucial part of mathematics - look at knot theory for example (Matt Parker made a video on that recently).

There's a difference between applying valid criticism and building an entire online persona around it.
You can only talk so much about one issue. If you want to stay relevant as a content creator, you need to create engagement.
And I think that's where she's losing touch with reality

[dire igloo]

Funding is flawed.
Publication bias is a thing.

But pointing fingers at expensive projects doesn't help, there have to be solutions.
And from the few things I've seen of her, she's not trying to produce solutions.
She wants to whine about academia bad and boast her credentials to discredit critics

<@&387163995947270144> scam spam

[thin trout]

<@&387163995947270144> #off-topic-tech message

[twin dew]

<@&387163995947270144> Spam

[dire igloo]

XD

[thin trout]

😄

[dire igloo]

This is just beautiful

[thin trout]

https://youtu.be/86yFyYJiPds?t=12

[dire igloo]

The issue in this discussion is simple:
She has valid points. She brings up valid criticism.
But she does it in a way that moreso resembles a right wing grifter than a scientist.
And beyond the broad picture of her criticism, she just doesn't really provide much input

[thin trout]

I take more issue with the stuff that she doesn't have valid points about. She uses her platform as a vehicle to pick out 5% of the evidence, ignore the 95% of higher quality evidence which contradicts her and pretend that it doesn't exist, then paint her per-formed bigoted view as a legitimate viewpoint to the detriment of real people (e.g. supporting policies that strip vulnerable groups of essential human rights)

it's nazi stuff and it's indefensible

[twin dew]

And so easy to lie by omission when grifting like that.

[soft bloom]

Yes, redundancy is a thing...
I was starting to write message about different... Flavours of science, but must have not finished and it went forgotten.
I don't say that we need to get rid of width or depth of research. Problem is that it seems like we got no breakthroughs in... Decades.
Correct me if Icm wrong, but the last big leap was Higgs bozon, which took decades to prove. And before that we got... Lazers.
I mean as truly groundbreaking discoveries.
Ofcourse we got a lot of research, theories, consumer tech since that made it's impact. I am not dismissing that.
I say (and found her words align with) that without breakthroughs we wonct get far. Eventually we will run out of redundant research that isn't just abstract math making up new rules and seeing what sticks (or likely not).
It sort of... Become a beaten road to pick a popular safe topic, apply for funds and do your job. Which from the very beginning has very small probability to make breakthrough because it's "safe".
This reminds of how devastated I was when heard about what projects students did for bachelour and masters degree. There's not even a hint of mystery, they are copy pasta. Heck half of them are translations done with google translate!

[pure karma]

teamwork makes the dreamwork! or however the saying goes

[dire igloo]

Didn't we detect gravitational waves only very recently?

If your standard for "breakthrough" is too high, no advancement will be enough

[pure karma]

iv been crashing out over javascript for the past like 6 hours

its been fun

[dire igloo]

And for that last part:
That's either anecdotal evidence which holds no value or it's a failure of good education which should be called out as that

Plus: the only thesis for which you are required to provide something new is a PhD

[dire igloo]

I was trying to stick with particle and astro physics there.

We've built things like light powered gears to build micro mechanisms.
Chemists have found a new way of synthesizing Flourine based oxidizers which drastically increased production capacity

And neuroscience has made some crazy advancements too, both scientific and socially.

Going from "autistic children are bad" to "we need autistic people for a productive society".
Going from "chronically sad? get drunk" to SSRIs

[night girder]

It's simple. You have AI ciritics. You have AI lovers. And then there is all the rest.

AI critics -> Terminator will become reality.

AI lovers -> AI will solve climate change.

All the rest -> 🤷‍♂️

Sabine Hossenfelder has been a AI critic from asfar as I remember. That's why I found the recent YT (I posted it a few days ago I believe) video about the AI bubble refreshring because it looked at all the angles and divided them into "camps". So it was trying very hard to be nuanced.

And "nuanced content" is a fucking rarity on YT.

[dire igloo]

I've never seen anything of her in regards to AI.
It was all just about ivory tower academia

[night girder]

🤨

[dire igloo]

This stuff

[night girder]

Right ...

Ok, I agree with what you said there. 😄

Just saw someone say Sabine is infamous for pseudo science.

[dire igloo]

Yeah, she on the brink of deranging from appropriate criticism of modern academia into an science-denialism grifter

[night girder]

But didn't you say; "It's not because you don't understand what scientist say, it's wrong?"

Oh, it was about actual research.

[dire igloo]

Sounds too ambiguous for something I said.
But if you mean smth like "just because you don't understand it doesn't mean it's wrong" then yeah, that sounds like something I could've said

[night girder]

Well, I don't know I haven't watched Sabine in a while tbh. She predicted the end of AI ages ago. Still waiting for it lol.

[night girder]

imo, the majority of "this research is useless" criticism comes from a lack of understanding and not because it's actually useless

I was just thinking, maybe Sabine is far smarter than us, and we don't see what she sees.

Something a lot of people with high IQ can struggle with. They see things a lot don't. Or they connect dots a lot can't etc etc.

[dire igloo]

I'm yet to hear her mention the late stage capitalism world we're living in as part of her criticism

[night girder]

but like I said, I haven't watched for a while.

Lately, I can't even find any science I actually like ... (on YT)

Even Neil Degrass is being weird.

[dire igloo]

Read through the reddit thread.
Plenty of people from various fields that chime in á la "I work in quantum computing, her takes on it show she didn't even understand the fundamentals"

[thin trout]

Veritasium and nilered are my favs

[dire igloo]

Alpha Phoenix is nice for electronics
Extractions & Ire for backyard chemistry
The Thought Emporium for bioengineering

[night girder]

Ok, Veritasium I still hold in high regareds. The other one I don't know enough.

[dire igloo]

Nilered is cool too, I hope we'll see his air series soon now that he can mass produce LN2

[thin trout]

He mostly screws with expensive but interesting stuff 😛

https://www.youtube.com/watch?v=GmwaJnj6pfY

[dire igloo]

NileRed definitely worth a watch then if you're into chemistry

[night girder]

Oh, him! Now I see him in his "lab suit" 😄

But like, we need more science in this day and age. Or listen to more scientist?

[thin trout]

https://www.youtube.com/watch?v=d6Pcp944sRI

https://www.youtube.com/watch?v=RGw6fXprV9U

[night girder]

What about Steve Mould?

[thin trout]

He's fine

[night girder]

https://www.youtube.com/watch?v=NtZaP8VMv0c

NASA likes him 😊

They set up video call to recreate his experiment in space.

Funny part; it was teams I believe Just regular teams.

Honestly. I still think it's magic. The chain in the cup gaining velocity upwards.

[dire igloo]

More of an educational entertainer and communicator than a scientist, but yeah, definitely also a good content creator

[night girder]

Wait. Ok. So what makes Veri a scientist? And not Steve? What makes Nilered a scientist?

[dire igloo]

Tho I guess with the chain fountain stuff he also made it into science

[night girder]

Do you need to have a degree in science? Do you need to have scientific papers published? Or is the act of science enough to call on scientist.

[thin trout]

science is technically more about performing experiments to gain information via the scientific method, but they are all generally more science communicators / explorers

[dire igloo]

nothing imo, at least for Veritasium
NileRed experiments with theoretically possible chemical transitions (toilet paper moonshine), Steve Mould had the chain fountain stuff

[thin trout]

Cool Worlds is a good science channel by the more pure definition

[dire igloo]

I'd say The Thought Emporium is an example of a channel that leans closer into the science side of things

Ex&F is mainly about recreating old chemistry with limited means

Styropyro could be considered a science explorer in the context of radio frequency physics

[night girder]

Something Fireworker said made me think the following;

1. "Should science be responsible for the translation to the common folk? So everyone understands their science. Or is that the job of other people, teachers, journalists, documentaries, ..."
2. "Is it even needed for science to be translated to the common folk? Ivory Tower Science sounds bad ... but not everyone cares. Some only care for the results ... not the journey."

Because what I like about Veri etc. Is how he makes it so ... degestable for my brain 😄

[dire igloo]

Veritasium sparked the speed of electricity discussion, but imo the actually relevant science there was made by Alpha Phoenix

[night girder]

Oh I remember that one 😄

[dire igloo]

1. Both. Scientists need to make their work accessible to educators. Educators need to make science accessible to the public. Good scientists can bridge that gap themselves.
2. Absolutely yes. It is very much necessary. The amounts of anti-intellectualism, science-denialism and misinformation in the modern world is frightening. Educating people to know how things work is the only way to ensure things keep working.

Just look at Milo Rossi's videos about Philip Zieba or Shane Vibes Truth

[night girder]

I think I understand you. Atlough I have to be loose with the word "accessible" you use 😄

To me accessible is; the paper is published on X website for everyone to read.

But if that paper is written in science language only 1% can figure out ... that would defeat the purpose.

[dire igloo]

But if you don't understand the paper, the science within that paper remains inaccessible

[night girder]

Not really, you can access it.

Not read it. Or understand.

[dire igloo]

The paper? Yes.
The science? No.

[night girder]

"their work accessible"

But like I said, I know what you mean. Just translation/interpretation on my part 😄

And yeah, I think I agree. Maybe not with point 2 fully.

[dire igloo]

If you stand before a 5m wall, the area behind it is inaccessible.
If you're an experienced climber or brought help in form of a ladder, it becomes accessible.

But just because nobody is stopping you from climbing over that wall doesn't mean the lands behind it are accessible

[dire igloo]

My understanding of access refers to the research/work itself, yeah.
Not whatever means they chose to document it

[dire igloo]

Nobody needs to know everything, I agree with that.
But in the modern world, science and society have become too far detached from one another.

The only reason why conspiracies about academia can flourish is because science feels inaccessible

[night girder]

I can understand a society needs time to adjust to new found science.
But it's scary to see "old" and well established science being denied or challenged with ... mostly nothing. Just words.

[dire igloo]

That's exactly the point.
People grow detached from science.
And it not only affects new research but it also discredits established consensus

[night girder]

And honestly, I have no clue how to fix that 😄

First thing most will think is; education

but then there is also transition from science books, to science on radio, to science on TV, to science on the internet.

And with each transition, the volume went up, but the quality/gate-keeping went down?

[dire igloo]

The issue is money.
Communicators will move to whatever medium and methodology makes them money.
And with capitalism becoming more extreme over the last decades, so did public media

[bold tapir]

what the vr set should i get?

considering buying it in late spring

currently i think quest 3/3s

[mental oriole]

I can't watch him because he repeats "though" so much it gets annoying to listen/ watch to.

[dire igloo]

That might explain where I get my tendency from to use that word

[soft bloom]

just like being excited too often will make you blind for breakthrough.
those are extremes ofcourse, but there's difference.
light powered gears are not a breakthrough, that's more of combining 2 things we know about intoa single thing in a new application.
breakthrough is when we discover something what shakes our previous understanding of how world works, like wave-like behaviour of photons and electrons, and this spurs a miriad of new discoveries to follow, extending our understanding of how world functions.

[dire igloo]

So yeah, the speed at which breakthroughs occur will stagnate.
The more we know about the world, the less new stuff there is to discover.

The closest we have to that is a disproven conjecture in knot theory:
https://youtu.be/Dx7f-nGohVc

And maybe the Einstein tile

[tribal kraken]

Printer working nonstop. 2nd mask already painted. Just below 50 hours to print this woth 0.6mm nozzle and 0.4mm layers, about max the hot end can handle/melt. Going to spend some time to make welding mask headband fit to it.

[soft bloom]

the less new stuff there is to discover.
what makes you assume that it will work liek that?
19-20 centuries where quite dense with science, before that we got some progress, medieval times were quite calm for europe, and before that it's not like there are many discoveries recorded.

[dire igloo]

The amount of information about the universe is finite.
The closer we get to knowing everything, the fewer things there are to still discover

I'm not saying that we're anywhere close to knowing everything.
But it's also futile to assume that discovery speed will remain high indefinitely

[faint turret]

"The amount of information is finite"
What's this assumption based on? Is it kinda like "mass is finite, so there's finite info about it"?

[dire igloo]

The assumption that if knowledge is infinite, there is always something you don't know.
So even if you knew every law of the universe and the exact location and velocity of every particle everywhere, you could still be caught by surprise

[soft bloom]

I was writing large message talking about how interval between breakthroughs was quite uneven through history and that according to established timeline of human history there were housands of years without much scientific progress.
(I am just skeptical of that timeline, that's all. I just have much less confidence in history when it goes further then couple hundred years)

[dire igloo]

Issue is: the further back you go, the less documentation you'll find.
And let's not forget the role that politics plays in controlling knowledge: like the Catholic Church dictating a geocentric model of the universe during most of the middle ages

The gaps in the time line don't necessarily mean that there was no scientific advancements.
It could be that scientific research was suppressed in order to maintain political power.
It could be that documentation of that science was either lost or became unintelligible.
It could also have been erased from historic records by someone trying to maintain political power

[soft bloom]

those are exactly reasons why.
too many factors that can distort our perception of ancient times

makes me wonder if someone made ancient pcb to make calculation if we were to ever know about that. it's not like archeologists would expect to find anything like this - it might look like intricate piece of stone...
oh wait, modern kids already don't know about floppy disks

[young canopy]

There were a lot of things that archeologists first thought were decorative but later turned out to be calculation tools similar to slide rulers.  Basically mechanical computers.

While static electricity and unsteady electrical current had likely been discovered time and again to some degree, through history.  I highly doubt the steady and reliable DC current had been created that would be required for electronics.

That said, contrary to popular belief, mechanical computers date back to the 18th century, and turring capable ones (meaning they have the operations necessary to run full procedural programs) date back to at least 19th century.   Considering the Greek mechanical contraptions lost to time (mostly because of being made from wood), I am certain they likely had at least mechanical calculators if not full turing capable mechanical computers.

There are legends that the Greeks made highly accurate mechanical astrolabs and planetariums.  Those require extremely advanced mechanical computers that are capable of trigonometry.

[dire igloo]

If there was PCB, there'd be residues.
If there was electric circuitry, it's likely the Romans would've adapted it

Absence of artifacts is one thing, presence of other artifacts is another.
It's unlikely that a room with a soot covered ceiling and torch holders on the walls was illuminated by electric lights

[dire igloo]

Also: there's technological evolution.
From wood fire to oil lamps to gas lamps to acetylene torches, etc
From chemical batteries to steam powered generators.
From manual motion to steam to electric motors.

[cyan crescent]

Will i need drivers or anything for a dual 10g network card on trunas?

[dire igloo]

So while you may not find artifacts for every step of the process, you will most likely find some from other steps.
And if there is zero documentation about any of those steps, it's likely that whatever you interpreted as the end result must've been something else.
Like the Dendera Lightbulb - there is no documentation of cables, circuitry, power generation or anything else you'd need to power that lightbulb.
So it's most likely not actually a lightbulb

[glossy glacier]

Linux handles drivers via the kernel so generally no
But certain NICs may not be supported, Google with your specific NIC in advance

[dire igloo]

I mean, technically drivers are always needed, it's just the question of whether you need to install them

[twin dew]

And for 10G network adapters, everyone has in-kernel drivers, as long as the kernel is new enough and the distro provider included them in the compile of the default kernel.

Also, need to check if you mean TrueNAS (based on Debian Linux), or TrueNAS Legacy/Core (Based on FreeBSD).

[night girder]

Anyone know of a log location for AMD adrenaline if it crashes? Can't seem to find it.

[twin dew]

Where current TrueNAS (Linux) 25.04 has kernel 6.12.
And TrueNAS Core (BSD) 13.0 is based on FreeBSD 13.0.

[night girder]

Checked;

C:\Program Files\AMD\atikmdag_dce.log
%appdata%/local/AMD/CN

need logs on the applicaiton itself, not performance. The application dissapears randomly and I want to know why.

A bit annoying, want to take screenshot. Doesn't work. Go to desktop, look at tray, don't see icon, manually booting AMD adrenalin.

And DDU was done a week or two ago. So it's pretty much clean.

[twin dew]

Where those TrueNAS Core (BSD) versions are EOL and only receiving security updates.

[night girder]

This is wild, I created a 3rd person project in UnrealEngine.

To test out the water you know? All left on default.

Fucker goes into first person 🤦‍♂️

Good job epic 👍

[mental oriole]

[night girder]

Was playing a game made by a single developer and it started to itch again. And it was in Unreal Engine 5.

[twin dew]

I wouldn't call that first person.
Third person with the camera at 0,0,0 (inside the head).

[night girder]

No shit... 🤣 But the fact is, camera is miles behind character, so it's something else doing it 🤷‍♂️

You can see where camera is etc.

[night girder]

Also, if you view from inside the head, that's pretty much first person view. Only difference here is that it's not from the eyes and you see mesh of head.

[twin dew]

Ok.

[soft bloom]

FPV is usually done from torso

you know, like why in SF you could get into weird spider model with head squished or how else to describe it

[night girder]

I don't know? 🤣 I do find some tutorials doing it from the head.

But I haven't checked them out. Focusing on other things.

[dire igloo]

You'll know why it's torso once you try to add a gun with hipfire

[soft bloom]

experts™ made them

also preview might be intentionally like this cause that's what audience would expect

[pure karma]

<@&387163995947270144> 👍

[languid gulch]

welp, corporate greed killed off my new mmo

[jagged snow]

Oh?

That was fast\

[languid gulch]

billion dollar game run by one of the most profitable companies on earth

still not greedy enough apparently

[fallen cradle]

How 2 use Xp in '25, step 1. Cry

[mental oriole]

Looking at dat ass

[soft bloom]

that's what third person view is for, isn't it?

[mental oriole]

Yup :))))

[thin trout]

https://old.reddit.com/r/gaming/comments/1oikpw5/battlefield_6s_odd_exposure_made_this_guy_seem/?sort=top

[soft bloom]

Contrary to the control group, continual pre-training of 4 LLMs on the junk dataset causes non-trivial declines (Hedges' g>0.3) on reasoning, long-context understanding, safety, and inflating "dark traits" (e.g., psychopathy, narcissism). The gradual mixtures of junk and control datasets also yield dose-response cognition decay: for example, under M1, ARC-Challenge with Chain Of Thoughts drops 74.9 → 57.2 and RULER-CWE 84.4 → 52.3 as junk ratio rises from 0% to 100%.
^ 1/4 of overview from:
https://llm-brain-rot.github.io/
discovered from: https://www.youtube.com/watch?v=XhJYNSssqrA

github.io is not allowed? sad

[tough owl]

I gotta train an LLM soley in brainrot to see what it does

[soft bloom]

it will yap about 6 7

[night girder]

You want more?

was thinking about learning blender too ... 🤯