[FRIAM] "SSRN-id3978095.pdf" was shared with you

[Steve Smith]

This was stuck in my mailtool, waiting for me to complete a closing 
sentence and hit send...

On 1/22/24 12:46 PM, glen wrote:

> Words matter: how ecologists discuss managed and non-managed bees and 
> birds
> https://link.springer.com/article/10.1007/s11192-022-04620-2

Words do matter, acutely in the circumstance that modern humanity 
(hominids since spoken/written language?) live (almost entirely) in a 
co-created intersubjective (religious, political, economic) reality.

re: "loaded language"... This is what I'm finding fascinating about the 
LLMs I've engaged (mostly GPT3/4 and Bard), is how *well* they seem to 
infer the context of any discussion I'm trying to have with them.  They 
can *patently* get it overtly wrong, especially (I infer) if there are 
some rule-based guardrails built into the system I'm interacting with.  
In human discourse, this is what the *worst* of PC (usually on the left) 
and "dog whistles" and other "coded language" on the right generates.

>
> I'm continuing this thread because I really want to classify types of 
> ad hominem. The most obvious bifurcation is human vs. source ... so 
> against the human versus against the source. E.g. I've been using 
> https://ground.news. And this morning, they had some 100% right (as in 
> political right, not correct) article where all the sources were Mixed 
> or Low Factuality. I found myself trusting the Mixed Factuality Fox 
> News more than any of the others. So ... familiarity? I guess?

I've also been using groundnews.com and finding it mildly interesting 
but in the spirit of GI:GO it feels as if the "meta-signal" they 
provide, while useful, is somewhat limited by the quality of the 
journalism they are working with.  I suspect that oldSkool Journos (e.g. 
Tom Johnson here) are extremely frustrated with the *average* quality of 
the work put out there as well as the curatorial/editorial 
distortions/biases that are generally available.   I'd be interested in 
others' experience with the various news-bias detectors/raters/recommenders.

>
> Re loaded language and the birds and the bees: I don't think it's 
> possible to remove the loading from loaded language. The best you can 
> do is be aggressively transparent about your loading. And that's what 
> triggers my "ad hominem". It's a reaction to closed or obfuscated 
> loading. E.g. when some older white dude at the pub insists on using 
> politically incorrect language that makes the kids cringe, you have to 
> do a little analysis and modeling of the speaker. Are they doing it 
> 'cause they're just too stupid? 'Cause it stirs up the kids? 'Cause it 
> makes them feel "free"? Etc. It's less *against* the person and more 
> of an attempt to infer their loading.
My experience with being in some sense the "old white dude" (pick your 
period, pick your ephithet, it is probably more to the point: "presumed 
to have some specific bit of power over the other who is confronting or 
commentarying on me") is that I had to shift from 'trying to be liked" 
to "trying to be understood" to "trying to understand myself as embedded 
in a larger system" to "trying to understand myself, full-stop".   The 
escalation from liked to understood to self-understanding doesn't likely 
help anyone else appreciate (or tolerate) me, but it seems to be a step 
above self-gratifying rationalization?
>
> On 1/8/24 11:06, Nicholas Thompson wrote:
>> I am particularly grateful for the ad hominem stuff.
>