Alexander Schmidt-Lebuhn
@anschmidtlebuhn.bsky.social
Botanist, taxonomist, phylogeneticist.
Pinned
Phylogenomics and genetic admixture of the invasive fireweed complex (Senecio inaequidens–Senecio madagascariensis) at a global scale - Biological Invasions
The southern African Senecio inaequidens–S. madagascariensis complex, (‘fireweed complex’), contains several species that have established as weeds outside of their native ranges. Senecio madagascarie...
doi.org
We have expanded our target capture dataset of the Senecio inaequidens - S. madagascariensis complex to a global scale. More clarity on the areas of origin of invasive madagascariensis, but also raising new questions.
doi.org/10.1007/s105...
@biolinvasions.bsky.social
#asteraceae #compositae #weeds
doi.org/10.1007/s105...
@biolinvasions.bsky.social
#asteraceae #compositae #weeds
As there seems to be another round of that discourse happening among people I am following:
The fight over the definition of "AI" is unwinnable. Everything a computer does semi-autonomously is being called "AI", even scripted opponents in strategy games. The fight has been over for decades.
The fight over the definition of "AI" is unwinnable. Everything a computer does semi-autonomously is being called "AI", even scripted opponents in strategy games. The fight has been over for decades.
December 1, 2025 at 12:27 AM
As there seems to be another round of that discourse happening among people I am following:
The fight over the definition of "AI" is unwinnable. Everything a computer does semi-autonomously is being called "AI", even scripted opponents in strategy games. The fight has been over for decades.
The fight over the definition of "AI" is unwinnable. Everything a computer does semi-autonomously is being called "AI", even scripted opponents in strategy games. The fight has been over for decades.
This is bad, but it is also always fascinating merely to get a glimpse into how other areas of science handle their processes. I cannot recall ever receiving a peer review report for a conference submission. It is only a committee reading my abstract and confirming that it is accepted.
The enshitification of scientific publishing marches on. First arXiv had to clamp down on submissions thanks to a flood of AI-written papers.
Now ICLR authors say their peer reviews were churned out by AI. Reviews full of hallucinated content that didn’t exist and offering useless, vague feedback.
Now ICLR authors say their peer reviews were churned out by AI. Reviews full of hallucinated content that didn’t exist and offering useless, vague feedback.
Major AI conference flooded with peer reviews written fully by AI
Controversy has erupted after 21% of manuscript reviews for an international AI conference were found to be generated by artificial intelligence.
www.nature.com
November 30, 2025 at 10:49 PM
This is bad, but it is also always fascinating merely to get a glimpse into how other areas of science handle their processes. I cannot recall ever receiving a peer review report for a conference submission. It is only a committee reading my abstract and confirming that it is accepted.
I find posts like these funny too, but in any serious analysis, an explanation that involves smartphones/social media as a load-bearing part would have to grapple with e.g. the 1920s-30s and the late Roman Republic. We humans just aren't as wise as the we think we are.
That begs the question: why has everyone lost their damn minds?
And the answer I think is Phone x COVID x Neoliberalism = everyone is insane.
And the answer I think is Phone x COVID x Neoliberalism = everyone is insane.
November 30, 2025 at 10:37 PM
I find posts like these funny too, but in any serious analysis, an explanation that involves smartphones/social media as a load-bearing part would have to grapple with e.g. the 1920s-30s and the late Roman Republic. We humans just aren't as wise as the we think we are.
This sounds like a great study, but some of the quotes they put into this write-up are ... strange?
"it apparently doesn't take a large brain to get into the air"
Well done, you have now figured out what everybody else figured out as a child when they saw saw a mosquito for the first time.
"it apparently doesn't take a large brain to get into the air"
Well done, you have now figured out what everybody else figured out as a child when they saw saw a mosquito for the first time.
November 29, 2025 at 10:48 PM
This sounds like a great study, but some of the quotes they put into this write-up are ... strange?
"it apparently doesn't take a large brain to get into the air"
Well done, you have now figured out what everybody else figured out as a child when they saw saw a mosquito for the first time.
"it apparently doesn't take a large brain to get into the air"
Well done, you have now figured out what everybody else figured out as a child when they saw saw a mosquito for the first time.
There should be some affordance for making mistakes or missing non-obvious problems. But seriously, nobody who lets stuff like this slip through should be allowed to claim on their CV that they are an editor. Did they even look at the manuscript?
"Runctitiononal features"? "Medical fymblal"? "1 Tol Line storee"? This gets worse the longer you look at it. But it's got to be good, because it was published in Nature Scientific Reports last week: www.nature.com/articles/s41... h/t @asa.tsbalans.se
November 28, 2025 at 1:46 AM
There should be some affordance for making mistakes or missing non-obvious problems. But seriously, nobody who lets stuff like this slip through should be allowed to claim on their CV that they are an editor. Did they even look at the manuscript?
And here I am, a fool, carefully reading through an 81-page long manuscript and flagging where the authors aren't consistently using past tense.
"Runctitiononal features"? "Medical fymblal"? "1 Tol Line storee"? This gets worse the longer you look at it. But it's got to be good, because it was published in Nature Scientific Reports last week: www.nature.com/articles/s41... h/t @asa.tsbalans.se
November 28, 2025 at 1:40 AM
And here I am, a fool, carefully reading through an 81-page long manuscript and flagging where the authors aren't consistently using past tense.
Two thoughts:
1. Nobody claims it is the panacea; it is a heuristic. We wouldn't want to remove speed limits either because they do not ensure zero accidents.
2. What matters is good faith peer review. Of course some for-profit journals only pretend to do it, and they should be shunned.
1. Nobody claims it is the panacea; it is a heuristic. We wouldn't want to remove speed limits either because they do not ensure zero accidents.
2. What matters is good faith peer review. Of course some for-profit journals only pretend to do it, and they should be shunned.
Opponents of preprints may crow "told you so". But peer review isn't the panacea (plenty of papermill/AI slop in journals) and the type of filtering/screening needed is not the same as the j̶u̶d̶g̶m̶e̶n̶t̶ ̶o̶f̶ ̶v̶a̶l̶i̶d̶i̶t̶y̶ hierarchical ranking of articles journals provide 3/3
November 27, 2025 at 9:16 PM
Two thoughts:
1. Nobody claims it is the panacea; it is a heuristic. We wouldn't want to remove speed limits either because they do not ensure zero accidents.
2. What matters is good faith peer review. Of course some for-profit journals only pretend to do it, and they should be shunned.
1. Nobody claims it is the panacea; it is a heuristic. We wouldn't want to remove speed limits either because they do not ensure zero accidents.
2. What matters is good faith peer review. Of course some for-profit journals only pretend to do it, and they should be shunned.
Maybe there should be some kind of "five indicators that you are in an abusive relationship" flashcard collection that is just handed out to everybody when they turn sixteen.
on top of everything else, nuzzi has the sensibilities of your most embarrassing friend. a guy alternates between saying he wants her to have his children and ghosting her, and she's wringing her hands trying to figure out if they're really in love
November 27, 2025 at 2:56 AM
Maybe there should be some kind of "five indicators that you are in an abusive relationship" flashcard collection that is just handed out to everybody when they turn sixteen.
Reposted by Alexander Schmidt-Lebuhn
“I wish someone at Google would remember scholar exists!”
…and the monkey paw curls.
…and the monkey paw curls.
November 25, 2025 at 3:08 PM
“I wish someone at Google would remember scholar exists!”
…and the monkey paw curls.
…and the monkey paw curls.
The singularitarians and AI bros simply do not understand how science works. Hypotheses actually have to be tested with empirical data and experiments. Faster thinking doesn't do anything there.
They could just as well invest into Hollywood movies to increase timber production. Cargo cult stuff.
They could just as well invest into Hollywood movies to increase timber production. Cargo cult stuff.
Ok so.
"The Genesis Mission will build an integrated AI platform to harness Federal scientific datasets — the world’s largest collection ... — to train scientific foundation models and create AI agents to test new hypotheses, automate research workflows, and accelerate scientific breakthroughs."
"The Genesis Mission will build an integrated AI platform to harness Federal scientific datasets — the world’s largest collection ... — to train scientific foundation models and create AI agents to test new hypotheses, automate research workflows, and accelerate scientific breakthroughs."
"the president has directed the Department of Energy to launch an AI project called the genesis mission" sounds like something out of a technothriller
www.whitehouse.gov/presidential...
www.whitehouse.gov/presidential...
November 25, 2025 at 11:07 AM
The singularitarians and AI bros simply do not understand how science works. Hypotheses actually have to be tested with empirical data and experiments. Faster thinking doesn't do anything there.
They could just as well invest into Hollywood movies to increase timber production. Cargo cult stuff.
They could just as well invest into Hollywood movies to increase timber production. Cargo cult stuff.
Reposted by Alexander Schmidt-Lebuhn
We should have listened when the modems screamed at us.
July 22, 2025 at 2:30 PM
We should have listened when the modems screamed at us.
Capitulescence of a Cassinia (#Asteraceae: Gnaphalieae) at the Australian National Botanic Gardens.
November 25, 2025 at 2:38 AM
Capitulescence of a Cassinia (#Asteraceae: Gnaphalieae) at the Australian National Botanic Gardens.
It is kind of funny how regularly the autocomplete suggestions for a search on any software or operating system include "disable AI".
November 25, 2025 at 12:09 AM
It is kind of funny how regularly the autocomplete suggestions for a search on any software or operating system include "disable AI".
The fee would have to come from somewhere. Assume USD400, as frequently suggested, a rejection rate of 50% after one round of with two reviewers, and accepted papers on average needing only a second round with one reviewer. That means a publication fee of currently USD2000 would have to be USD4000.
I think we need to pay reviewers.
25 no response out of 30 requests to review!
Drives me mad this selfish behaviour, which I also see as a journal editor. And it's almost always established folk with secure jobs, not least those endlessly brandishing their right-on-ness on social media
Drives me mad this selfish behaviour, which I also see as a journal editor. And it's almost always established folk with secure jobs, not least those endlessly brandishing their right-on-ness on social media
November 24, 2025 at 8:55 PM
The fee would have to come from somewhere. Assume USD400, as frequently suggested, a rejection rate of 50% after one round of with two reviewers, and accepted papers on average needing only a second round with one reviewer. That means a publication fee of currently USD2000 would have to be USD4000.
If you don't buy new stuff even when your old stuff still works, the economy suffers. Between that and building the entire civilisation on non-renewable energy sources, people a thousand years from now will consider us all to have been collectively insane. But for most of us, it's just Monday.
November 24, 2025 at 3:20 AM
If you don't buy new stuff even when your old stuff still works, the economy suffers. Between that and building the entire civilisation on non-renewable energy sources, people a thousand years from now will consider us all to have been collectively insane. But for most of us, it's just Monday.
Nymphoides (Menyanthaceae), presumably N. montana, near Braidwood last week.
November 23, 2025 at 11:20 AM
Nymphoides (Menyanthaceae), presumably N. montana, near Braidwood last week.
This seems kind of obvious, once you take a step back and think about business models. But the entire bubble has not thinking about business models at its core. How many AI coding startups are just funneling queries to OpenAI or Anthropic, meaning that customers could go to the latter directly?
One thing I'd caution against those who say "what my company does can be done entirely with AI... I don't need to pay actual people"...
If that's true, then what are you building that your customers can't just use ChatGPT to get?
www.theguardian.com/education/20...
If that's true, then what are you building that your customers can't just use ChatGPT to get?
www.theguardian.com/education/20...
‘We could have asked ChatGPT’: students fight back over course taught by AI
Staffordshire students say signs material was AI-generated included suspicious file names and rogue voiceover accent
www.theguardian.com
November 20, 2025 at 8:13 PM
This seems kind of obvious, once you take a step back and think about business models. But the entire bubble has not thinking about business models at its core. How many AI coding startups are just funneling queries to OpenAI or Anthropic, meaning that customers could go to the latter directly?
Great and wide-ranging read. A different, and less 'mystical', way of putting it:
Maybe AGI would be possible but only by also replicating the trade-offs of the biological brain such as cognitive biases, brittle ego, cancer, and mortality. Maybe you can't eat your electronic cake and have it too.
Maybe AGI would be possible but only by also replicating the trade-offs of the biological brain such as cognitive biases, brittle ego, cancer, and mortality. Maybe you can't eat your electronic cake and have it too.
In which I articulate the case that AGI is impossible and machines just plain cannot think. www.eruditorumpress.com/blog/on-inco...
On Incomputable Language: An Essay on AI
An incidental consequence of having written a book on tech-fascism and the so-called rationalist movement is that I find myself periodically queried for my thoughts on artificial intelligence. On the ...
www.eruditorumpress.com
November 18, 2025 at 8:17 PM
Great and wide-ranging read. A different, and less 'mystical', way of putting it:
Maybe AGI would be possible but only by also replicating the trade-offs of the biological brain such as cognitive biases, brittle ego, cancer, and mortality. Maybe you can't eat your electronic cake and have it too.
Maybe AGI would be possible but only by also replicating the trade-offs of the biological brain such as cognitive biases, brittle ego, cancer, and mortality. Maybe you can't eat your electronic cake and have it too.
Reposted by Alexander Schmidt-Lebuhn
What's great about this is that one of them will absolutely nail it, and you're impressed, and then it refreshes after a minute and the same model will be laughably bad, and this is 100% my professional experience with LLMs.
Hahaha! Er. Someone has used AI to make analogue clocks with the live time.
Analogue clocks are really hard.
AI: spectacularly confident. Can be both subtly and spectacularly wrong. This is a lovely illustration of that.
clocks.brianmoore.com
Analogue clocks are really hard.
AI: spectacularly confident. Can be both subtly and spectacularly wrong. This is a lovely illustration of that.
clocks.brianmoore.com
AI World Clocks
The current time as rendered by 9 different AI models. By Brian Moore.
clocks.brianmoore.com
November 17, 2025 at 12:27 PM
What's great about this is that one of them will absolutely nail it, and you're impressed, and then it refreshes after a minute and the same model will be laughably bad, and this is 100% my professional experience with LLMs.
It would be nice if bad things were regulated, and if regulations against bad things were enforced, but apparently "bad things not happening" is not something most voters find that important these days.
Collecting bubbles right now.
* AI bubble
* Wall Street bitcoin
* Private equity bubble
* adding the Buy Now Pay Later bubble, which is ... largely untracked!!
techcrunch.com/2025/11/16/b...
this feeds to my theory that the economy is screwed so we get *multiple* financial scams all at once
* AI bubble
* Wall Street bitcoin
* Private equity bubble
* adding the Buy Now Pay Later bubble, which is ... largely untracked!!
techcrunch.com/2025/11/16/b...
this feeds to my theory that the economy is screwed so we get *multiple* financial scams all at once
'Buy Now, Pay Later' is expanding fast, and that should worry everyone | TechCrunch
As Morris watches his BNPL investments from the other side of the table, he seems to understand the warning signs better than most.
techcrunch.com
November 17, 2025 at 12:06 PM
It would be nice if bad things were regulated, and if regulations against bad things were enforced, but apparently "bad things not happening" is not something most voters find that important these days.
The Youtube ad selection algorithm in its wisdom decides to serve me, a cyclist Luddite researcher, SUVs, AI-generated vampire fiction, crypto trading apps, and payment systems for small businesses.
I am half-convinced that advertising is a fraud committed against advertisers.
I am half-convinced that advertising is a fraud committed against advertisers.
Also, despite having endless information on my buying history for well over a decade, Amazon is still like: "based on your past purchases of Fimo clay, printer ink and reading glasses, we think you'd like this Hallowe'en banana costume, penis shaped ice cube tray & Jeffrey Archer's autobiography." 🙄
November 16, 2025 at 8:30 PM
The Youtube ad selection algorithm in its wisdom decides to serve me, a cyclist Luddite researcher, SUVs, AI-generated vampire fiction, crypto trading apps, and payment systems for small businesses.
I am half-convinced that advertising is a fraud committed against advertisers.
I am half-convinced that advertising is a fraud committed against advertisers.
I hope, but am not optimistic, that people understand the difference between
(a) I could not do my job without this database.
and
(b) I could not do my job without the hallucinating, plagiarising, text and image generator.
(a) I could not do my job without this database.
and
(b) I could not do my job without the hallucinating, plagiarising, text and image generator.
November 16, 2025 at 8:21 PM
I hope, but am not optimistic, that people understand the difference between
(a) I could not do my job without this database.
and
(b) I could not do my job without the hallucinating, plagiarising, text and image generator.
(a) I could not do my job without this database.
and
(b) I could not do my job without the hallucinating, plagiarising, text and image generator.
Reposted by Alexander Schmidt-Lebuhn
If I ever believed this, you couldn't drag this admission from me
November 16, 2025 at 12:40 PM
If I ever believed this, you couldn't drag this admission from me
Look at that, the genus I did my PhD about also occurs in Chile!
dx.doi.org/10.4067/s071...
#minthostachys
dx.doi.org/10.4067/s071...
#minthostachys
dx.doi.org
November 16, 2025 at 4:24 AM
Look at that, the genus I did my PhD about also occurs in Chile!
dx.doi.org/10.4067/s071...
#minthostachys
dx.doi.org/10.4067/s071...
#minthostachys
What gets me about alt text is that unless the image contains text or is a visual gag, it generally shouldn't even be necessary, because the post itself should be informative. Like, if you post an image of a bug, building, machine, or person, just put its name into the actual post under it.
another reason i love alt text is because a lot of y'all post photos of pop culture people and as someone who wasn't allowed to watch anything but SESAME STREET and CARE BEARS and shit until i was like 10... i have no idea who those people are until i read the alt text.
November 14, 2025 at 9:52 PM
What gets me about alt text is that unless the image contains text or is a visual gag, it generally shouldn't even be necessary, because the post itself should be informative. Like, if you post an image of a bug, building, machine, or person, just put its name into the actual post under it.