Even if you disregard, y’know everything, LLM fans are still fundamentally boring.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur ugh, the number of times I have to close the video these days after "and here's the interesting part".
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
A little issue. (I’m not saying there are no “LLM fanatics” or ideologues)
But the war-of-religion-inducing word here is “fan”
You seem to have no trouble distinguishing:
- faith
- frequency of use
- enthusiast
- advocate
- supporter
- proponent
- die-hard
- expert
- intermediate
- novice
- ceo
...So why collapse all of that into “fan”?
That word smuggles in too much. My point of viewI like the reasonable distance taken by #codeberg and #mathematician in the domain of automatic theorem proving
They do not totally condemn usage while at the same time warning and even distancing themselves from embracing the technology mindlessly
I think their position is super acute as it bypass the questioning about intellectual property: open source and mathematical knowledge are less prone to money induced issues and what could be labelled as "steal" or "ruthless copy"
Some mathematicians have asked for a moratorium on datacenters-led AI
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur that is the most damning thing about Hank Green saying that his LLM tells were not scripted
-
@baldur that is the most damning thing about Hank Green saying that his LLM tells were not scripted
@qim Absolutely.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
That feeling when LLM fans reply to the above post and prove exactly the point you were trying to make.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur has anyone written an article about LLMs as a zombie infection
-
A little issue. (I’m not saying there are no “LLM fanatics” or ideologues)
But the war-of-religion-inducing word here is “fan”
You seem to have no trouble distinguishing:
- faith
- frequency of use
- enthusiast
- advocate
- supporter
- proponent
- die-hard
- expert
- intermediate
- novice
- ceo
...So why collapse all of that into “fan”?
That word smuggles in too much. My point of viewI like the reasonable distance taken by #codeberg and #mathematician in the domain of automatic theorem proving
They do not totally condemn usage while at the same time warning and even distancing themselves from embracing the technology mindlessly
I think their position is super acute as it bypass the questioning about intellectual property: open source and mathematical knowledge are less prone to money induced issues and what could be labelled as "steal" or "ruthless copy"
Some mathematicians have asked for a moratorium on datacenters-led AI
@stphrolland @baldur Sealion alert.
-
@baldur has anyone written an article about LLMs as a zombie infection
@robinsyl Not that I know of.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur The revolution will be homogeneous.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur I don’t do YouTube, but I’ve come to recognize the diction and the cadences of AI narration from the continuous drone coming up from the basement, where my husband is half-watching history videos while snoozing.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur I'm starting to more view the novel anti-ai movements in terms of herseiology, maleficarum and the apocalyptic traditions.
-
@baldur has anyone written an article about LLMs as a zombie infection
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur I don't think this is necessarily true. I do however think that continued interaction with LLMs whether willing or not will have an impact and shape people's perception and way of thinking/behaving. Much as we're susceptible to repeated patterns outside of LLMs.
I have someone close to me that I do not in any way consider boring.
They are a vibrant and hardworking person, but I've still heard them regurgitate the gospel of the tech bros "You're going to fall behind if you don't.".
-
@baldur has anyone written an article about LLMs as a zombie infection
”Dawn of the Data Centers”
”28 Tokens Later”
”The CEO with All the Gifts”
”Night of the Vibing Dead”
Not sure if I’d watch any of these, though.
-
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur From a cultural perspective, LLMs makes everything uniform, in the worst way possible. As the output is all very uniform, it influences its users to be the same. There was research not too long ago that showed that LLMs affected the users' thoughts, and that this didn't change even if people were aware of this phenomena.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur I mean I thought douchy tech bros all sounded the same even before LLMs.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur At a recent software conference, I was bracing for one of the keynotes to be especially bad... and it *did* fill out my entire "AI tropes that induce despair or rage" bingo card. But it was actually much more "meh" than "argh", because the talk was so clearly written and/or edited by LLM. It was a bland echo of someone trying to copy a brash opinionated persona, made even funnier because a decade ago this speaker would've easily hit the tone he seemed to be aiming for.
-
Even if you disregard, y’know everything, LLM fans are still fundamentally boring. They have the same writing and design style, their illustrations look the same, and their ideas are all predictable
They’re even starting to sound the same on video, even when unmediated by a model
@baldur now I really want to see a study following people as they start to use LLMs more, to see if their everyday speech starts to have less variety, maybe of vocabulary or sentence construction.
-
@baldur I don't think this is necessarily true. I do however think that continued interaction with LLMs whether willing or not will have an impact and shape people's perception and way of thinking/behaving. Much as we're susceptible to repeated patterns outside of LLMs.
I have someone close to me that I do not in any way consider boring.
They are a vibrant and hardworking person, but I've still heard them regurgitate the gospel of the tech bros "You're going to fall behind if you don't.".
@baldur Here's to me realising that this could be applied to a number of my familiars in hindsight. I was writing about someone specifically close to me, but I know several people this could describe.
All of them vibrant and kind, but unwilling to accept what they are doing and participating in is harmful. If there's anything they do have in common it's that they want to "see the positive in everything" and are 'doers' rather than analytically inclined.
