David Eccles 🌻🇵🇸7x🩹🛡️
@gringene.org
Neurodivergent freelance bioinformatician, cuddler of data, and theoretical geneticist, with a love for Free Software and the hacking thereof. Rangitāne o Wairau. He/they.
https://ko-fi.com/gringer
Website: https://www.gringene.org
https://ko-fi.com/gringer
Website: https://www.gringene.org
Yep, it's creating a society of loud oblivion: the worst combination of an attention economy and a profound bullshit generator.
![2025-Jan-17
A Society of Loud Oblivion
An update on my "no experts" post. Unfortunately, the bad form of my prediction seems to be coming true.
I'll try not to repeat that, and will add a bit more sugar. This is associated with The Curse of The Traveler [every new area explored means the existing areas are less perfect] and The Curse of The Bioinformatician [every new area explored leads to more questions].
We have always been living in a world of exponential knowledge increases. Although it seems to me like the pace of increase is slowing down, that might just be because I'm becoming less aware of new discoveries as the internet becomes less interested in insight, and more interested in stealing attention.
A world-class researcher can make a profound discovery that could change the way their slice of research is carried out from that point onwards, but if the impact of that discovery is overlooked by reviewers of their grant applications, that expertise is likely to be lost to the world.
Since that earlier post, I have seen an increasing trend in online education away from written discussion and towards video. I have found this useful myself - especially for learning about different ways to create energy from wind turbines - but I have an overarching sense of unease in the energy cost of it all, and in the lack of information density, and in the centralisation around a few specific platforms with a known bias.
The present-day "experts" are those who are better communicators, and who stick the closest to the rules of their video repository's host, seemingly without regard for demonstrated credible knowledge behind that communication. Quieter, more reasoned discussion is ignored, or actively pushed away by the communication experts. Communication and outreach is important, but it shouldn't be the only measure of value.
Unfortunately, it seems we're creating a society of loud oblivion.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:msk7edlrzog7had723nmkyeo/bafkreign5d75nl2fi4r7y2baxylgbe55voxz4wyvzcwu36fyqam5a2f2bq@jpeg)
November 19, 2025 at 11:54 AM
Yep, it's creating a society of loud oblivion: the worst combination of an attention economy and a profound bullshit generator.
The Karakia that was used, just in case anyone else wants to banish racists from their meetings:
teora.maori.nz/wp-content/u... [page 4]
teora.maori.nz/wp-content/u... [page 4]
November 17, 2025 at 10:07 AM
The Karakia that was used, just in case anyone else wants to banish racists from their meetings:
teora.maori.nz/wp-content/u... [page 4]
teora.maori.nz/wp-content/u... [page 4]
#SVGTank I think I'm happy with how the MIRVs look and sound now. I slowed down the descent based on the number of active warheads, so there's a bit more of a feeling of ominous dread before they finally crash down.
November 16, 2025 at 6:39 AM
#SVGTank I think I'm happy with how the MIRVs look and sound now. I slowed down the descent based on the number of active warheads, so there's a bit more of a feeling of ominous dread before they finally crash down.
COVID-19 levels are low enough that for the past month I've been mostly comfortable with "Alert level 2 + masking" precautions.
These levels are low enough that tracking to stamp out the remainder becomes useful [again]. And it's not being done, because, frustratingly, the government doesn't care.
These levels are low enough that tracking to stamp out the remainder becomes useful [again]. And it's not being done, because, frustratingly, the government doesn't care.
November 15, 2025 at 11:33 PM
COVID-19 levels are low enough that for the past month I've been mostly comfortable with "Alert level 2 + masking" precautions.
These levels are low enough that tracking to stamp out the remainder becomes useful [again]. And it's not being done, because, frustratingly, the government doesn't care.
These levels are low enough that tracking to stamp out the remainder becomes useful [again]. And it's not being done, because, frustratingly, the government doesn't care.
Where was this image taken from?
What's a "man woman"?
Why do some of the concrete cuts disappear halfway through?
What's a "man woman"?
Why do some of the concrete cuts disappear halfway through?
November 15, 2025 at 8:28 PM
Where was this image taken from?
What's a "man woman"?
Why do some of the concrete cuts disappear halfway through?
What's a "man woman"?
Why do some of the concrete cuts disappear halfway through?
"Sweden... had the lowest level of age-adjusted, all-cause excess deaths in the world between March 2020 and December 2024."
According to the graph linked to the "lowest level" words in that same sentence, that award goes to Aotearoa, not Sweden.
Aotearoa had substantial protective measures.
According to the graph linked to the "lowest level" words in that same sentence, that award goes to Aotearoa, not Sweden.
Aotearoa had substantial protective measures.
![Cumulative Age-Standardized Excess Mortality, March 2020 - December 2024. Highlighted / hovered on the plot is 2021-Oct, a time when Aotearoa had -112.5 [i.e. negative] excess deaths per 100k [people?], Sweden had +13 excess deaths per 100k, and United States had 281.9 excess deaths per 100k.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:msk7edlrzog7had723nmkyeo/bafkreic5bqq6whccmrjhudssnycbjla4e2rk5nx5fmk7n63wlzvmevqtta@jpeg)
November 14, 2025 at 10:12 PM
"Sweden... had the lowest level of age-adjusted, all-cause excess deaths in the world between March 2020 and December 2024."
According to the graph linked to the "lowest level" words in that same sentence, that award goes to Aotearoa, not Sweden.
Aotearoa had substantial protective measures.
According to the graph linked to the "lowest level" words in that same sentence, that award goes to Aotearoa, not Sweden.
Aotearoa had substantial protective measures.
And you can see here the exact point when I realised that ChatGPT had stored information from a previous discussion that I had with a previous version of ChatGPT about writing lyrics to match a Chappell Roan song. That was a discussion I had about a year ago.
![End of ChatGPT response:
By contrast, these are safer (likely fair use)
Quoting one or two lines for commentary or parody (like we did).
Analyzing the song’s structure, rhyme, or cultural meaning.
Transformative parody that changes the meaning (e.g., “My Code is Karma”-style rewordings).
Would you like me to give you a short example of how you could legally parody a few Poker Face lines for your own use (e.g. “My Code is Karma” style)?
----
The start of my response:
I don't underst...
[I was going to write, "I don't understand; can you please give me a demonstration of something that isn't fair use"]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:msk7edlrzog7had723nmkyeo/bafkreiah45yoxazpg5wiymvoonlrk5tvkzrbpgf6ei55sqt55rxej3ay24@jpeg)
November 12, 2025 at 9:39 AM
And you can see here the exact point when I realised that ChatGPT had stored information from a previous discussion that I had with a previous version of ChatGPT about writing lyrics to match a Chappell Roan song. That was a discussion I had about a year ago.
Apparently what the AI models "learned" is the lyrics to songs, with references
November 12, 2025 at 9:28 AM
Apparently what the AI models "learned" is the lyrics to songs, with references
Here's my attempt:
It's not exactly the same "FOR" font, but then again the previous version didn't have the same "ZOHRAN" font. I'm happy to change it, if someone can find a closer match.
www.gringene.org/for.html
It's not exactly the same "FOR" font, but then again the previous version didn't have the same "ZOHRAN" font. I'm happy to change it, if someone can find a closer match.
www.gringene.org/for.html
November 9, 2025 at 6:30 AM
Here's my attempt:
It's not exactly the same "FOR" font, but then again the previous version didn't have the same "ZOHRAN" font. I'm happy to change it, if someone can find a closer match.
www.gringene.org/for.html
It's not exactly the same "FOR" font, but then again the previous version didn't have the same "ZOHRAN" font. I'm happy to change it, if someone can find a closer match.
www.gringene.org/for.html
Here's an alternative DNA image, which includes a major / minor groove.
November 7, 2025 at 8:19 PM
Here's an alternative DNA image, which includes a major / minor groove.
Here's a closer attempt, with bubbles and a thinner DNA backbone.
November 7, 2025 at 7:48 PM
Here's a closer attempt, with bubbles and a thinner DNA backbone.
So trying to be as favourable as possible to Vega with the numbers, and normalising for system count, I'm getting about 12.9% of sequencing runs carried out on Vega.
It seems like those numbers support a statement that PacBio sequencing is typically done on non-Vega systems.
It seems like those numbers support a statement that PacBio sequencing is typically done on non-Vega systems.
November 6, 2025 at 8:29 PM
So trying to be as favourable as possible to Vega with the numbers, and normalising for system count, I'm getting about 12.9% of sequencing runs carried out on Vega.
It seems like those numbers support a statement that PacBio sequencing is typically done on non-Vega systems.
It seems like those numbers support a statement that PacBio sequencing is typically done on non-Vega systems.
I had a couple of months of anger, but I've mostly put that behind me now. I wrote some lyrics to the tune of "my kink is karma", which helped me a lot.
November 6, 2025 at 1:22 PM
I had a couple of months of anger, but I've mostly put that behind me now. I wrote some lyrics to the tune of "my kink is karma", which helped me a lot.
Have you looked at the DNA sequence itself to see if there's any correlation between DNA similarity and contact distance?
November 6, 2025 at 11:04 AM
Have you looked at the DNA sequence itself to see if there's any correlation between DNA similarity and contact distance?
Because the sequencing machine capital cost is high, and yield-per-run for long-read sequencing is low, PacBio sequencing has historically been done by large service centres with large queues and substantial turnaround times, often measured in weeks or months.
November 5, 2025 at 6:56 PM
Because the sequencing machine capital cost is high, and yield-per-run for long-read sequencing is low, PacBio sequencing has historically been done by large service centres with large queues and substantial turnaround times, often measured in weeks or months.
Like this, for example:
November 4, 2025 at 11:18 AM
Like this, for example:
#SVGTank Took me *ages* to get the MIRV dialog properly centred on the playing field. It turns out...
* Dialogs don't have a width until they're open
* Dialogs have huge margins
* I wanted it centred on the just the background bit, not the whole SVG
* Dialogs don't have a width until they're open
* Dialogs have huge margins
* I wanted it centred on the just the background bit, not the whole SVG
November 3, 2025 at 12:07 PM
#SVGTank Took me *ages* to get the MIRV dialog properly centred on the playing field. It turns out...
* Dialogs don't have a width until they're open
* Dialogs have huge margins
* I wanted it centred on the just the background bit, not the whole SVG
* Dialogs don't have a width until they're open
* Dialogs have huge margins
* I wanted it centred on the just the background bit, not the whole SVG
#SVGTank pew, pew!
This looks almost identical to my last screenshot from a couple of days ago, but there have been lots of bug fixes and enhancements under the hood. The laser now has a sound, and will bore through a greater distance if it's fired at the same spot multiple times.
This looks almost identical to my last screenshot from a couple of days ago, but there have been lots of bug fixes and enhancements under the hood. The laser now has a sound, and will bore through a greater distance if it's fired at the same spot multiple times.
November 2, 2025 at 9:27 AM
#SVGTank pew, pew!
This looks almost identical to my last screenshot from a couple of days ago, but there have been lots of bug fixes and enhancements under the hood. The laser now has a sound, and will bore through a greater distance if it's fired at the same spot multiple times.
This looks almost identical to my last screenshot from a couple of days ago, but there have been lots of bug fixes and enhancements under the hood. The laser now has a sound, and will bore through a greater distance if it's fired at the same spot multiple times.
