Isn’t that how it works? They just grab the first study they see and don’t even read it

Has she done the thing where she answers a question with a screenshot of an AI overview yet?

I’m enjoying this conversation while stuck at the airport 😂

That is...not wrong.

Found it. Not exactly a question, but tried to prove my supposed misunderstanding by doing that. bsky.app/profile/quac...

Is there anything wrong in that text?

To start, it relies on a false premise you input - that I was misconflating the biology term hermaphrodism. But, it does helpfully demonstrate that people exist who can produce both gametes, even if they cannot successfully gestate, so again proves your definition of sex means it's bimodal.

No one produces both gametes and can play both sexual roles. Unlike biological hermaphrodites like some snails.

That was never your assertion. Your assertion was sex is defined by which size gametes someone had. The fact that people exist with both gametes demonstrates that your definition of sex is that sex is bimodal. You can try to move the goalposts all day long. We all know you lost. Badly.

Your own text contradicted you, dumbass

It was written by an LLM. As Abe Lincoln said, "When you go to an LLM for answers, you have lost."

Was there anything wrong in the text?

Oh, so we’re being blissfully ignorant. Cool.

I mean, if you can’t see how that text contradicts your not-so-finely crafted one-line narrative, might need your vision checked

It’s true, I saw him say it.

Me too! I didn't know you were there too!

I was hanging in the back with Elvis and Bob Marley

Oh that's why I didn't see you. I was up front with Hendrix and Freddie.

It’s LLM output. It has no truth value whatsoever.

Is it correct? What is wrong in the text?

It conflates actual analysis from the human brain with inherently flawed algorithmic summation but you go, quackie.

So very dishonest of you.

Oh but that’s been the game all along with him

Why should I bother analyzing text created by a nonsense machine?

I adopt this text as my own. And will defend it.

Just as you've defended all of your claims, right?

You haven't successfully defended your premise that sex is binary and not bimodal so I have my doubts you can do it here LOL

Same question.

It makes quite good sense. I think this time it’s probably accurate. But the burden here is not on you to refute it. It is on the person who provides the text to back it up with citations to the literature (which funnily enough you get for free if you quote a scientific paper!)

Well, too, the falseness comes from the prompt given. This was output based on a prompt assuming someone was misunderstanding biological hermaphrodism and conflating it with humans who have DSDs.

Thanks! I noticed that, but didn’t think about it. It’s worth thinking twice about.

Sure, it might. But why bother even looking into it? The sources it links to, maybe, but I can’t get those from a screenshot.

Even if you could get them, it’s still bogus. The thing is that I typically want to see the list of references for the source of the quote (or paraphrase), and then the list of references for each citation in the quote, if I’m going to make a reliability or suitability for purpose judgment. 1/

The text is correct, in proving you wrong and showing your love of stepping on rakes

“…humans with this condition have a range of variations in their sexual anatomy and often do not have perfect sets of male and female organs, nor are they necessarily able to procreate with both…”

Yes, a lot actually. Particular when applied to YOUR previously stated positions regarding sex determination. That's the thing, you keep taking pieces from various areas - one off sentences or statements that "agree" with you and try to stitch them together into an entire thought.

The problem you have is that, when placed in their proper context, they DON'T agree with you, often in fact CONTRADICT you, or at BEST are not relevant to your position. But you stitch them together and claim it all adds up to "the science says [you're] right" But that's not a valid argument

As for your "is there anything wrong with that statement." It's literally impossible to know because you cannot evaluate the "correctness" of an AI answer without at least two things: 1. The full text of the prompt 2. Access to the sources it's drawing from

You do not need to know these things in order to know if the statement is correct or not. FFS. If I got the LLM to state "Lyon is the Capital City of France" you would would not need to know prompts etc to know if it correct. You could look it up independently.

If you’re sharing something as an assertion of fact, sure. We could research the statement to figure out whether it’s BS or not. If you’re citing something as an authority, and you really must use an LLM, we’d need the prompt too.

I am not citing the LLM as an authority. I fully adopt the output as my words. Am I correct or not?

An LLM is just advanced pattern recognition. The patterns aren’t necessarily correct. That’s why LLMs keep going all racist and such. There’s racism in the patterns used to create the model. Racism in, racism out. BS in, BS out too.

If you could look it up independently, then the LLM output wasn't necessary or relevant in the first place.

It was a good way of explaining to you. But my mistake was expecting you to read it and engage in good faith.

Yes, actually, we do. Sorry that you don't understand how questions or LLMs work in addition to not understanding gender is bimodal, but that's your own ignorance.

I adopt the LLM words as my own. Am I correct in saying there are two meanings to the term "hermaphrodite"? And we are not discussing "gender" - whatever that is.

Nonsense. You could find your own sources to verify it. My point is that the term hermaphrodite is used as an equivocation. Am I right or wrong?

You're flatly wrong.

Then you simply do not understand. The term is used to describe how a species reproduces and is used for types if rare development disorders. Two different meanings that are conflated by people who do not know what they are talking about.

How are we in like week 8 and still don't have an answer to 'if gamete production is dispositive, what about people who produce both or neither?'

You still haven’t responded to my answer of your thought experiment.

The "you could find" line is funny because (a) given the hypothesis is that the quack is repeating nonsense, it follows you really won't find verification (can't verify falsehoods), and (b) producing anything that contradicts El Quacko results in assertions that you found "the wrong" evidence.

'"hermaphrodite" conflates a natural, non-human reproductive phenomenon (having both male and female reproductive parts) with a human medical condition, true hermaphroditism' -- yeah I hate when a term conflates hermaphroditic animals and people with 'true hermaphroditism'

No human can reproduce using both reproductive roles. That is what defines biological hermaphroditism. You are mixing up non-evolved, rare medical conditions of development with an evolved reproductive strategy in some creatures like molluscs.

Such obvious dishonesty. I never claimed, nor did anyone else, that a human can reproduce using both reproductive modes. Your adopted definition of sex has nothing to do with reproduction. It has to do with gamete size. So you again can't prove a binary and have to resort to blatant lies.

Developmental conditions that aren't environmental are necessarily evolved.

It could easily be straight MAKING up it's sources.

If a LLM described water as h2o would you doubt its sources? You are free to argue that this argument is right or wrong since I fully adopt it as my own.

Thing is, the LLM part is superfluous at best and it's a Gish gallop at worst. It's not that it's right or wrong, it's that LLM is extremely unreliable to the point that it's reckless, even fraudulent, to rely on it.

Which is to say, we should consider the source of this information to be… you?

In many cases AI answers will come from AI generated work, and AI can't tell the difference.

Or the AI has trained itself on Reddit sh*tposts and spews out the half-masticated results.

lollll, every fuckin time

This is text I am willing to stand behind. That is why I posted it. Do you want to say it is wrong and why?

I already told you what was wrong and why and how it proves sex is bimodal.

And you were incorrect and the paper never says what you wish it to say

LOL It does, in fact, say what I think it says. But it's okay to admit you misread something or didn't fully understand what you read.

So again. Let’s start with the basics. Where does it say there are more than two sexes?

Yes, the basics. As you define sex, there are more than two of them because you define sex based on the size of gametes. So at least four. Male, female, a sex comprised of people who do not have proper gonads to produce gametes, and an additional sex of people who can produce both kinds of gametes.

That's not what bimodal means.

What sex is someone who produces no gametes?

My God, you really are a broken record

You don’t get to stomp around saying “you’re incorrect” and not say why. Are you five years old?

He doesn’t care. In his mind, he’s right and that’s all that matters.

It's just very entertaining how in trying to prove he's right he keeps proving he's wrong.

Oh it is. I’ve been just rapt here

Why should I believe it’s anything other than nonsense? After all, an LLM wrote it.

It also doesn’t enforce a single thing they claim. Words can have multiple meanings, both lay and scientific.

It’s clear that you don’t or are willfully trying not to understand. Because doing so would undermine the same line you’ve been repeating over and over again.

Even with an actual scientist, here in the thread, you still insist on your perfect little binary view

🤣🤣🤣

I'm pretty sure. Let me find it.

As a scientist, I feel we in the community have to take partial responsibility for this. We started writing abstracts for SEO by simplifying the language and generalizing, almost like we were writing the teaser for a novel, and people are reading that and then deciding that's the whole book.

Maybe it’s worse than when I got my PhD 40 years ago, but clickbait abstracts preceded the existence of clicks by many decades. And it’s not clear to me that it’s actually worse, on net, because it’s also symptom of a vast expansion in the number of people working in science. 1/

There’s also the “desk drawer problem”, that people are strongly biased to presenting and accepting for publication “significant” results (with a few exceptions like pharma based in government regulation), while “no such effect” results tend to end up in a stack in a desk drawer. 2/2

I think that it was more of a problem in the soft sciences in the past, and has now transitioned to the soft sciences. It wasn't too long ago that reading an abstract for a physics or biochemistry article required a degree in physics or biochemistry to decipher. But agree re "desk drawer problem"

*transferred to the hard sciences. [Inserts IV with caffeine into neck]

LOL. It’s still true as originally poasted!

I work as a PCT in an oncology unit and I’ve wondered if that’s a thing

Particularly when the abstract is the only thing available if you don’t have (fiendishly expensive) journal access.

This is true. But whoever decided that turning abstracts into clickbait articles was a good idea needs to get smacked upside the head. It's not generating more journal subscribers, but LESS. Because people wrongly assume that the abstract is summarizing the report.

This is, to me, the duality of SEO. It's needed to get the traffic. But it necessarily is incomplete and seen by a much wider audience without enough knowledge or education to understand the difference between an abstract and the actual outcome.

And in the process makes it very easy to strip out the context and make it seem like the report is saying something it's not. "Evaluating the efficacy of NIOSH rated masks in an office environment" becomes "SCIENTIST at Harvard doubts masks are safe!" We saw this with climate change science.

The simplest and most effective solution would be to publish both the abstract AND the conclusion outside of the paywall. Because often times the abstract is just our summarization of the process, and the conclusion is a summarization of the results. It's still a narrative, but more complete.

Sadly it works too, people tend to prefer simplicity even if any further review would contradict their point

To give them all benefit of the doubt, we place scientific journals behind such paywalls that unless you're a university or a research department, you can't afford to read them. I can't afford to read the journals I'm published in without subsidy. If the author can't, how can the laypeople?

This has been a frustration for me. I have three rare diseases & a rare manifestation of failure of a medical device (the device often fails but how it failed for me was rare). It's near impossible to stay up to date. Even worse is it's near impossible for my doctors to do so. For the same reasons.

Especially since things like journal subscriptions are the first expenses that hospitals and labs will cut to save on costs. And I don't care how much you make, those journals will end up making your budget look like a dril candle tweet.