neRok

Thanks for sharing. I've been questioning the validity of my own ADHD "diagnosis", but I had only been considering things with regards to dopamine and the "reward-system", and not the executive-function and mental "optimisations" side of things like he explained. This finally gives reason to all my traits/quirks/problems!

I called it my "diagnosis", because even the psychiatrist that prescribed me didn't seem convinced, but he's also a bit old and thus maybe out of date (ie, only considering the traditional "ADD" style symptoms). So I've been on the drugs for a year now, and up to a BIG dose, but besides some very minor concentration and motivation improvements, the only worthwhile improvement was not being tired in the afternoons (I regularly used to nap on the train on the way home from work, which was at only 4pm! Sometimes from 3->4pm I would even struggle to keep my eyes open whilst at my desk!). So when the presenter said he had tiredness problems too, that was further confirmation for me (of a problem being fixed). Because I had been wondering if "the opposite" was happening, in that my improvement was simply the "stimulants" keeping me awake (ie, just being a "tweaker"), which would imply that the tiredness was normal (and to further confuse things - I have reviewed the traits reported by ancestrys dna test, and one of them says that I am very likely to take naps!).

Unfortunately I doubt this revelation will change anything for me, and that's because I've got decades of "optimisations" baked-in. For example, socialising is a big problem. Thinking back, and by 10 years old I was known to be "shy" - but recently I have come to consider this as "learned" behaviour (based upon a few memories). This new explanation about executive-function and "optimisations" is an even better explanation. Long story short - socialising is hard, and failures are bad, so I've optimised away the failures by not talking much! It's exactly like the example the presenter gave of Test Driven Development @ 23:00;

  > ...suddenly there'll be 500 broken tests and you're like where did that
  > come from what's going on and it turns out there's actually nothing wrong with what you've done to the code but you
  > have just broken all of the mock setup of all of the tests and this is something which
  > I find unbearable almost literally unbearable because I'm end
  > up I end up in this situation where I'm put in a position where either I go and
  > fix all of those tests which in my mind are obviously pointless and which is mind-numbingly boring or I have to have
  > the discussion in the pr review of why the code coverage has gone down on the CI
  > server and it drives me nuts and yes I will stop ranting now on that.

I'm part way through watching this now, but have already noticed that Steve has not understood what Nvidia was talking about regarding "securing budget";

    > [6:46] in order to continue doing these types of interviews
    > [6:48] and for them to quote secure budget we  
    > [6:50] had to cover things a certain way in the  
    > [6:53] reviews that were totally unrelated to  
    > [6:55] the people we were even interviewing

The budget Nvidia is referring to is for their employees time, because they are paying for that employee to be interviewed. In my experience in biggish companies with "aggressive" cost-codes/time-sheets systems is that "project staff" are a cost that need all their hours attributed to a project or similar, otherwise the projects KPI's don't look good. So these thermal and latency engineers are predominately hired for "project" roles, and so they would be assigned to work on developing new products, or supporting existing products.

What isn't in that "project scope" is spending many hours preparing for and then doing interviews. That's something that suits "marketing", and so that department/function would be responsible for paying for those work-hours. And so then it makes sense that those marketing people want value for money, and hence want all their talking points included as part of the "deal" for them to fund the employee to do the interviews.

I work in engineering/CAD sector and have had a similar experience. One time I was on a project and made a 3D animation to simulate a particular construction process that was being designed. Then the marketing people saw the animation and wanted a better looking version created that they could then use for various things. And so as was the way of the company - I gave the standard response of "sure, just give me a time code for xxx hours and I will get it done". Then they ask for a detailed breakdown of those hours with target dates etc, so then you have to respond with "sure, just give me a time code for 4 hours work and I will get that done".

[Going to give myself a self-reply here, but what-ev's. This is how I talk to chat-gpt, FYI]... So I happened to be shopping for a cheap used car recently, and we have these ~15 year old Ford SUV's in Aus that are comfortable, but heavy and thirsty. Also, they come in AWD and RWD versions. So I had a thought bubble about using an AWD "gearbox" in a RWD vehicle whilst connecting an electric motor to the AWD front "output", so that it could work as an assist. Here was my first question to chat-gpt about it;

  > I'm wondering if it would be beneficial to add an electric-assist motor to an existing petrol vehicle. There are some 2010 era SUV's that have relatively uneconomical petrol engines, which may be good candidates. That is because some of them are RWD, whilst some are AWD. The AWD gearbox and transfer case could be fitted to the RWD, leaving the transfers front "output" unconnected. Could an electric motor then be connected to this shaft, hence making it an input?

It gave a decent answer, but it was focused on the "front diff" and "front driveshaft" and stuff like that. It hadn't quite grasped what I was implying, although it knew what it was talking about! It brought up various things that I knew were relevant (the "domain knowledge" aspect), so I brought some of those things in my reply (like about the viscous coupling and torque split);

  > I mentioned the AWD gearbox+transfer into a RWD-only vehicle, thus keeping it RWD only. Thus both petrol+electric would be "driving" at the same time, but I imagine the electric would reduce the effort required from the petrol. The transfer case is a simple "differential" type, without any control or viscous couplings or anything - just simple gear ratio differences that normally torque-split 35% to the front and 65% to the rear. So I imagine the open-differential would handle the 2 different input speeds could "combine" to 1 output?

That was enough to "fix" its answer (see below). And IMO, it was a good answer!

I'm posting this because I read a thread on here yesterday/2-days-ago about people stuggling with their AI's context/conversation getting "poisoned" (their word). So whilst I don't use AI that much, I also haven't had issue with it, and maybe that's because of that way I converse with it?

---------

"Edit": Well, the conversation was too long for HN, so I put it here - https://gist.github.com/neRok00/53e97988e1a3e41f3a688a75fe3b...

I've recently started asking the free version of chat-gpt questions on how I might do various things, and it's working great for me - but also my questions come from a POV of having existing "domain knowledge".

So for example, I was mucking around with ffmpeg and mkv files, and instead of searching for the answer to my thought-bubble (which I doubt would have been "quick" or "productive" on google), I straight up asked it what I wanted to know;

  > are there any features for mkv files like what ffmpeg does when making mp4 files with the option `--movflags faststart`?

And it gave me a great answer!

  (...the answer happened to be based upon our prior conversation of av1 encoding, and so it told me about increasing the I-frame frequency).

Another example from today - I was trying to build mp4v2 but ran in to drama because I don't want to take the easy road and install all the programs needed to "build" (I've taken to doing my hobby-coding as if I'm on a corporate-PC without admin rights (windows)). I also don't know about "cmake" and stuff, but I went and downloaded the portable zip and moved the exe to my `%user-path%/tools/` folder, but it gave an error. I did a quick search, but the google results were grim, so I went to chat-gpt. I said;

  > I'm trying to build this project off github, but I don't have cmake installed because I can't, so I'm using a portable version. It's giving me this error though: [*error*]

And the aforementioned error was pretty generic, but chat-gpt still gave a fantastic response along the lines of;

  >  Ok, first off, you must not have all the files that cmake.exe needs in the same folder, so to fix do ..[stuff, including explicit powershell commands to set PATH variables, as I had told it I was using powershell before].
  >  And once cmake is fixed, you still need [this and that].
  >  For [this], and because you want portable, here's how to setup Ninja [...]
  >  For [that], and even though you said you dont want to install things, you might consider ..[MSVC instructions].
  >  If not, you can ..[mingw-w64 instructions].

I agree that the first example in the article is "bad"...

  fn frobnicate(walrus: Option<Walrus>)`)

but the rest makes no sense to me!

  // GOOD
  frobnicate_batch(walruses)
  // BAD
  for walrus in walruses {
    frobnicate(walrus)
  }

It doesn't follow through with the "GOOD" example though...

  fn frobnicate_batch(walruses)
    for walrus in walruses { frobnicate(walrus) }
  }

What did that achieve?

And the next example...

  // GOOD
  if condition {
    for walrus in walruses { walrus.frobnicate() }
  } else {
    for walrus in walruses { walrus.transmogrify() }
  }
  // BAD
  for walrus in walruses {
    if condition { walrus.frobnicate() }
    else { walrus.transmogrify() }
  }

What good is that when...

  walruses = get_5_closest_walruses()
  // "GOOD"
    if walruses.has_hungry() { feed_them_all() }
    else { dont_feed_any() }
  // "BAD"
    for walrus in walruses {
       if walrus.is_hungry() { feed() }
       else { dont_feed() }

I'm just a hobbyist and have some scriplets I've written to "improve" things on various websites. As part of that I needed an uninstall/undo feature - so with my "bush league" code I would do `window.mything = {something}`, and based upon previous dabblings with the likes of python/c#/go, I presumed I would be able to do `delete window.mything` and it would auto-magically call a function I would have written to do the work. So the new `[Symbol.dispose]()` feature/function would have done what I was looking for - but really, it;s not a big deal, because all that I actually had to do was write an interface spec'ing `{}.remove()` method, and call it where needed.

(This paragraph is getting off topic, but still... ) Below is my exact interface that I have in a .d.ts file. The reason for that file is because I like typed languages (ie TypeScript), but I don't want to install stuff like node-js for such simple things. So I realised vscode can/will check js files as ts on-the-go, so in a few spots (like this) I needed to "type" something - and then I found some posts about svelte source code using js-docs to type their code-base instead of typescript. So that's basically what I've done here...

  export global {
    interface Window {
        MyThing?: {remove: ()=>any}
    }
  }

So chances are that in the places you could use this feature, you've probably already got an "interface" for closing things when done (even if you haven't defined the interface in a type system).

From a technical/algorithmic POV - this doesn't sound particularly remarkable, because it sounds about the same as how modern CPU's "boost" themselves.

And actually, I think there is similar tech in "speaker systems" already. An "older" one that I've read about is from RCF (an old+big Italian company with systems ranging from desktop to festival size), and they call it [Bass Motion Control](https://www.rcf.it/en/art-9-series);

> The BMC method works by creating a complete map of the dynamic behavior of the woofer, to generate a custom algorithm that only limits over-excursions. This gives total freedom of signal reproduction to the transducer. When high-pass filters normally protect the woofer motion from becoming destructive but change the phase behavior, the new BMC algorithm breaks conventional rules.

Now I don't know how effective RCF's approach truly is, but another company that is doing "big-things" is Dirac. They released a blog post about a year ago titled [Boosting Audio System Sustainability with Dirac](https://www.dirac.com/blog/boosting-audio-system-sustainabil...), and there is a section called Enhancing performance with optimized components;

> By employing Long Short-Term Memory (LSTM) neural networks, we can make the driving force on the voice coil (the part of the speaker that turns electricity into sound) more consistent, improving the mechanical design and compensating for magnetic limitations.

> NLC adjusts the voice coil current to correct force factor irregularities (inconsistencies in the voice coil’s efficiency) without requiring complex mechanical measurements. In tests with an otherwise suboptimal driver, our technology reduced distortion by 10 dB, nearly matching the performance of a well-designed driver.

> I occasionally need the hauling capacity [of a] 1998 Ford Ranger

Then rent a suitable vehicle for the occasion.

- Your example Ford Ranger[1] seems to have towing capacity of 6,000lbs (~2700kg), and a payload capacity of 1,260lbs (~570kg).

- Compare that to the worst model Toyota Hilux[2], which has a payload of up to 1240kg, and can tow 2500kg. These can be rented for like $65 AUD per day (~$40 USD).

[1] https://www.kbb.com/ford/ranger/1998/specs/ [2] https://www.redbook.com.au/cars/details/2019-toyota-hilux-wo...

Banning its addition is a step beyond — but in Queensland, Australia for example, the state government no longer mandates its inclusion, and thus the local councils are able to set their own policy.

> While more than 90 per cent of Australians have access to fluoridated water, that figure is significantly lagging in the sunshine state, where local councils have ultimate authority over whether it is adopted.

> A decade after the Newman government handed responsibility for fluoridated drinking water to local governments, 51 out of 77 have opted out. That means about 28 per cent of Queenslanders do not have fluoridated drinking water

https://www.abc.net.au/news/2023-11-24/fluoride-dental-care-...

> Nothing is changed by creating a `const newStroke` and pushing it to the array.

Not true. If you look carefully you will see the `const` and everything after it is wrapped in curly braces, and those braces create a block...

> Traditionally (before ES6)... blocks with curly braces do not create scopes... [but] blocks are finally treated as scopes in ES6, but only if you declare variables with let or const. — https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guid...

So the curly braces with the use of const/let creates a new lexical scope, hence solving the issue. You can test it in the browser console to confirm;

  var stack=[], cmd="";
  cmd="success"; stack.push(()=>cmd);
  cmd="fail"; stack.push(()=>cmd);
  stack[0](); /* returns "fail" */

  var stack=[], cmd="";
  cmd="success"; {const c=cmd; stack.push(()=>c);}
  cmd="fail"; {const c=cmd; stack.push(()=>c);}
  stack[0](); /* returns "success" */

It's probably not a good idea to do it this way, because it's easily "overlooked". I would probably use a class or factory-function if I had the need.

I wonder if using structuredClone() has a performance penalty compared to a simple block-scoped closure though?

   let doFn, undoFn; {const newStroke = currentStroke; doFn = ()=>strokes.push(newStroke); undoFn = ()=>strokes.pop();}

LPG had its chance here in Australia in the 2000's, but it didn't work out [1]. Subsidised conversions were offered back then, which many people took up on their 80's+90's model cars, but it probably led to the early demise of many of those cars. By the late 2000's there were factory cars selling with LPG "gas" systems, but they weren't as good as the petrol variants [2]. By 2011 you could get "liquid injection" models, which were as good [3]. But like the first article states, people with "large" cars were shifting to SUV's and dual cabs by that stage, and a lot of them were 4cyl petrol/petrol-turbo/hybrid/turbo-diesel, so their fuel efficiency was comparatively decent.

Also in 2011, the government introduced fuel-tax on to LPG, which it had been exempt/discounted on previously. So the price went up, and all of a sudden it wasn't as economical. So that put it in to a death-spiral, especially when many servo's started getting rid of it due to low demand. Now people with LPG cars have "range anxiety" [4]. Where I live (Perth, Western Australia), you are basically stuck in the metro area if you are LPG only. It also means there's a ton of ~2008-2012 LPG Falcons for sale second-hand at good price, but no one wants to buy them.

Out of curiosity, I just went and looked up some data for where I live - the Perth metro (source [5]);

  Year  Month  Fuel  c/L avg  Servos  $cost / 100km for Ford Falcon sedan (UNO)
  -----------------------------------------------------------------------
  2010  Dec    LPG     69.4   216     @ 15L = $10.4  (gas model)
               ULP    128.0   298     @ 10L = $12.8  (petrol model)
  2019  Dec    LPG     89.0   147     @ 15L = $13.4  (gas model)
               LPG     89.0   147     @ 12L = $10.7  (LPi model)
               ULP    144.9   387     @ 10L = $14.5  (petrol model)
  2024  Dec    LPG    133.9    46     @ 12L = $16.1  (LPi model)
               ULP    172.7   436     @ 10L = $17.3  (petrol model)
               ULP    172.7   436     @  6L = $10.4  (Falcon Ecoboost 4cyl 2L turbo)
               Dsl    178.5   437     @  9L = $16.1  (Ford Territory SUV 2.7L TD)

[1] https://www.fleetcare.com.au/news-fleettorque/fuel-cards/wha... [2] https://www.smh.com.au/national/running-the-rule-over-lpg-op... [3] https://www.drive.com.au/reviews/ford-falcon-ecolpi-lpg-revi... [4] https://www.abc.net.au/news/2023-04-19/lpg-cars-disappearing... [5] https://www.fuelwatch.wa.gov.au/retail/historic

The article (which is from 2023 by the way) doesn't discuss the findings of the DNA analysis, so I went and looked up the original - https://doi.org/10.1038/s41586-023-06035-2

> with estimates of present-day human and faunal contamination both below 1%. Comparisons with present-day human populations²⁶ using ƒ3-statistics and D-statistics²⁷,²⁸ show high affinities to Native Americans (Extended Data Fig. 5). When projected into a principal component analysis with other ancient human individuals (Fig. 3c), DCP1 falls within a group of Ancient North Eurasian individuals from further east in Siberia, which includes the approximately 24 ka Mal’ta 1 and the approximately 17 ka Afontova Gora 3 individuals²⁹,³⁰. Both of these individuals are genetically closer to DCP1 than non-Ancient North Eurasian individuals when tested with D-statistics (Extended Data Fig. 6b), and all three show similar affinities to ancient Siberians and Native Americans with ƒ3-statistics and D-statistics

Apparently not. In Linus Tech Tips latest video, the Framework CEO says they talked to AMD about it and that one of their engineers ran some simulations and found the stability would degrade too far for it to be possible.

> No, this lawsuit is not about employee records. It is about customer records.

That's not what the article says...

> The plaintiffs include "unions and membership organizations representing current and former federal employees and federal student aid recipients and six military veterans who have received federal benefits or student loans,"

I wonder how much of this employee data is actually "personal", and not data relevant to their employment, which presumably isn't "private" from the perspective of the employer. So for example, I imagine their home address and birth date would be considered private; but their job title , primary place of work, start date, etc would not be "private data"...?

A basic speaker can't do a lot to improve "room acoustics" [1], particularly below the Schroeder frequency where room modes greatly affect the bass response. From my experience, it's the bass that needs fixing in room, because even a "perfect speaker" will get boomy/muddy at some frequencies (ie. reflections overlapping constructively), and thin/null at others (ie. reflections overlapping destructively). And if you aren't going to worry about "fixing the room", then there are already companies/products like Kali Audio IN-5 speakers [2] that have squeezed in some good performance and tech (eg. active DSP with coaxial driver) in to an affordable package.

There are ways to improve the bass situation, and one is by implementing "directional bass", aka a Cardioid array. This could be achieved with multiple individual speakers (add-on/upgrade existing system), or could be integrated in to 1 "speaker". Examples of the latter have been around for a few years - see Dutch & Dutch 8c or Kii Three. They are relatively expensive (which you could consider an early adopter tax), but affordable competition is starting to mature with speakers such as Mesanovic CDM65 [3].

There is another way to improve things too, and that is via "Active Room Treatment" [4] as Dirac calls it. Basically it uses excess capability in various speakers to "clean up" the audio of other speakers in the system by outputting "cancellation waves" (to cancel the problems). The results appear amazing, but they are taking their sweet time getting it released on to affordable equipment.

There's also "spatial audio" like Dolby Atmos that should/could work around room problems in a similar way to Dirac ART. So good speakers (like already exist) + ART + Atmos + AI "upscaled" 2 channel source music could be the final frontier? But that's just for "mechanical" sound reproduction. Maybe in the future I can just transmit the song straight in to my brain, bypassing my ears and the need for speakers entirely?!

[1] https://en.wikipedia.org/wiki/Room_acoustics [2] https://www.kaliaudio.com/independence [3] https://www.audiosciencereview.com/forum/index.php?threads/m... [4] https://www.dirac.com/live/dirac-live-active-room-treatment/

> Was that the case in the recent CrowdStrike debacle?

Nah, it still worked fine on my work laptop and allowed to boot in to safe mode, which also worked (didn't crash).

First I would make sure the brightness isn't too high, and then I would consider if the problem is text on screen. The LG OLED panels have WRGB pixel layout, so it doesn't work quite right with the likes of ClearType on Windows. Instead I use MacType on my PC with a modified profile and 117% zoom in windows, and now it's fine to me (the font was an issue when I first got my OLED).

My Gigabyte OLED monitor is 2 years old now and gets used for many hours on most days and shows no changes. I bugged out the time counter for a few months by putting the panel in to service mode, but still it is up to 4040 hours, or 5.5 hours per day for 2 years.

Your other post mentions the screen brightness dropping with white menus open, but that is not because of a burn in protection feature, but instead it's because of limits on the total power consumption of the panel (either because of power efficiency laws, or too much heat).

I also have aphantasia but can navigate just fine. I read a comment on here once from someone with aphantasia that couldn't navigate in the real world, but they worked in network admin and knew that layout just fine (knew where everything routed and what it connected to etc). I pondered that if they could remember cable routes, why don't they apply the same mental map to roads? Roads are fixed and have a beginning and end too! But they reckon they could not. So it's a strange problem.

There was a recent article on the seaweed: [Australian red meat industry says it doesn’t need to meet its self-imposed net zero target](https://www.theguardian.com/australia-news/2024/feb/08/austr...)

> This includes feeding cows seaweed based additives, but one of the longest commercial trials failed to meet hoped-for methane cuts and led to the animals eating less food. ... “The commercial viability of those [food additives] means they are not something likely to be widely adopted by the industry.”

So yes, business as usual. Nothing has changed.

> A 2022 CSIRO report commissioned by MLA found the industry emitted almost two-thirds the volume of greenhouse gases in 2020 compared with a 2005 baseline. But it said the reported reductions have been driven by a decrease in land clearing and an increase in forest regrowth, as recorded by Australia’s national carbon accounting system. An analysis by the University of Queensland said the NCAS may be grossly under-reporting land clearing rates.

I wondered this a long time ago too (if solar panels cause more heating by absorbing more), but actually they still reflect a fair portion of the suns energy. [This article](https://www.treehugger.com/ask-pablo-do-solar-panels-contrib...) says earths average reflection coefficient is ~0.35, and that cities are more in the range of ~0.1. Solar panels are ~0.3, so much better than many existing roof materials and of cities in general.

I have roller shutters on my east and west facing windows, and my north facing windows (this is in Australia) get no direct sunlight because they are shaded by my patio. So no direct sunlight is coming into my house through the windows. Still in summer after ~2pm the house heats up. I think overall my house is decently insulated from the sun because in winter my house is extremely cold (what little heat the sun could give doesn't get in to my house). Recently in early spring there were days when the outside air temp was hotter than in my house.

I think reducing the absorption of the roof and the walls would help, but even then you have still have a house sitting in 34C+ air temps for at least 10 hours per day, so it gets hot. But when I consider my dads house, which has 2m eaves all around, high ceilings, and massive roof space - it still gets hot. His house might feel nicer without a/c during early summer, but by mid+late summer it is hot too.

Maybe better insulated doors+windows would help? My house is double brick walls. Maybe the "outer layer" needs to be able to get hot, and then the "inner layer" insulated from the outer as much as possible. That's a lot of change though. Instead I just run the a/c all day "for free" off my solar panels.

It's just coming out of winter here in Perth, Western Australia. I have 6.6kw solar power and also a 15 year old solar hot water system with thermostat controlled electric booster (so not ancient, but also not the most advanced tech). On cold and cloudy days the booster turns on and draws far more power than my panels are generating at that moment, but even on nice days it turns on over night when there is no solar power. I saw on one particularly bad day that it turned on 3 times in 24h - 2x when dark - and drew over 1.5kWh each time, and I only used the hot water once in the morning. So I turned the booster off a month or 2 ago (can't remember when exactly, but yes it meant some days I couldn't have a hot shower), and on my recent 2-month power bill I had used 215kWh less than last year! (I also replaced my plasma TV with an OLED in that time, so that might have been responsible for a portion.) So whilst I get free hot water for ~8 months of the year, the other 4 months it uses a ton of electricity.

My sister recently got a new hot water system and I showed her some data that showed heat pump ones were better, so that's what she got. The data showed that whilst they use electricity during the summer, they use less during winter, and they end up ahead. But actually we can see on her power bill that when it kicks in on bad days, it too can overwhelm her solar generation at that moment.

Ultimately I think we need smarter connected systems to get the best efficiency. The booster in my solar hot water is just a basic element like a kettle. Surely it could be "throttled" (supplied with less watts) or cycled during the day so that it only heats when there is solar power available? If I have to have a slightly-cooler-than-hot shower first thing in the morning because I didn't want it boosting at 4am, then so be it.

I think the article made that point to counteract any questions of "who is the source, and how reliable are they on this topic?". Especially because Tim Flannery has already been the target of many "callouts" on Australian media, eg https://www.youtube.com/watch?v=7LzlEkxE-zs

It was just my first google result for "methane co2 equivalent", and so the number came from this link (am Australian) - https://www.cleanenergyregulator.gov.au/NGER/About-the-Natio...

You are right that there are different valid factors to consider, as can be seen here: https://en.wikipedia.org/wiki/Global_warming_potential

But also, 25x vs 100x is a 4x difference, which doesn't affect the comparison to coal that much (it's still single digit % "at best").

Just to confirm the numbers myself;

18,300kg methane per hour * 24 hours * 365 days = 160,308,000kg (~0.160 million metric tonnes). At 25x CO2 equivalent, that is 4,007,700,000kg (4 million metric tonnes).

This link about US electricity generation: https://www.eia.gov/tools/faqs/faq.php?id=74&t=11 shows 767 "million metric tons" from coal, or 1.55 "billion metric tons" from all sources.

4 / 767 ~= 0.5%, so in the ballpark of the parent comment. Also possibly the second link is ton (~1016kg) vs tonne (1000kg), further tweaking the numbers.

And just about the Permian basin, Wikipedia says it "accounts for 20% of US crude oil production and 7% of US dry natural gas production" - https://en.wikipedia.org/wiki/Permian_Basin_(North_America)#... So if all sites like this were measured, it might be more like 2.5% coal-use-equivalent?

I was recording TV well over 10 years ago with a Hauppauge TV Tuner card in my PC. I tried MythTV (on Linux), and Kodi and the like, but settled on MediaPortal (on Windows). I controlled the PC with a HTPC remote. I originally used MediaPortal's recording engine, but eventually switched to Argus TV, because it had smarter features (and integrated back into MediaPortal, so nothing was lost on the GUI side).

Anyway, the way most of these recording engines would work is basically that they would feed the recorded video through a "script" at the end, and so within that script you could send it to another program to find the ads. This was done with [ComSkip](https://www.team-mediaportal.com/wiki/display/MediaPortal1/C...) on MediaPortal and Argus TV. From memory, ComSkip would write a txt file or similar next to the video, and that was like a metadata file letting the playback program know where the ads were.

I'm not 100% sure how ComSkip worked internally, but there are many clues in the video stream it could use to find the ads (eg. often the static logo would disappear, or the teletext end, or the volume increase). It wasn't perfect, but it was good enough. I reckon with today's machine learning capabilities, such a program could be written to skip the ads perfectly 99% of the time. But I doubt anyone will, because even I haven't used my HTPC in this manner for over 5 years (I don't even have the TV connected to the aerial any more). Now my HTPC is just a fanless mini PC stick with Win11, a keyboard and mouse on the coffee table, and I stream through the web browser.

I think a lot of people that get wrapped up in flat earth stuff aren't the smartest, and so the problem starts in school when they are taught things like trigonometry, they often have a bad/dismissive attitude towards it (thinking "when would I ever use this in real life") or they just can't do it.

As an example, I have a friend that dropped out of high school relatively early to do a trade, and now he is keen about science and watches many videos on space and the like. One time he was theorising to me about how electricity in a wire makes a magnetic field, and that if you put it near a magnet then it would move. I said, that's what an electric motor is, lol. He probably would have learned that if he did better and stayed in school longer, but he wasn't interested back then.

But at the other end of the spectrum, the first person I really heard about flat earth from happened to be someone I went to highschool with. This was at our 10 year reunion (and was nearly 10 years ago now), and he had gone on to become a geologist, but had recently jumped onto flat earth. So, it's quite mind boggling to think about what was going on in his life that lead to that conclusion.

And 1 more thing about learning, I don't recall any schooling regarding 3D space, like how to think about it, approach it, etc. And then when talking about viewing/measuring space from the perspective of earth, you also have to consider your position on the globe, and how that relates to other objects. Even now I struggle with this aspect, and my job for nearly 20 years has involved 3D modelling on computers (although my work is 3D, it is all "planar" though, which is different to bodies in space).

There have been a few articles shared here in the last couple days about the Amish and their "herd immunity" event in May 2020 (seems like someone has an agenda). I think it is wishful thinking that they achieved it at that time (if they have at all). Tests done at the time and soon after were only returning positive results in the 60 and 70% ranges, which is less than required for herd immunity. Tellingly, they then seemed to fare no better when the biggest US wave hit: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8195242/

> Our results indicate the Amish/Mennonite excess death rates are similar to the national trends in the USA. The excess death rate for Amish/Mennonites spiked with a 125% increase in November 2020.

And this study doesn't even cover the current delta wave (stops Jan 2021).