Jan Kraemer
@camelusminimus.bsky.social
Anbieter feinster kognitiver Entropie.
Purveyer of finest cognitive entropy.
Currently being very confused by metaphors in computing, the objectives of ML, the discourse around AI, and the general impedance missmatch between meaning and maths.
He/Him
Purveyer of finest cognitive entropy.
Currently being very confused by metaphors in computing, the objectives of ML, the discourse around AI, and the general impedance missmatch between meaning and maths.
He/Him
If you used chatGPT at some point and want to be a kind soul, a (bachelor) student looks into issues of trust, experience and aspects of individual perception and could use your help:
Could you fill out this short (6-7 min) questionaire?
(german or english)
umfrage.hu-berlin.de/17701?src=1DSM
Could you fill out this short (6-7 min) questionaire?
(german or english)
umfrage.hu-berlin.de/17701?src=1DSM
February 8, 2024 at 10:40 AM
If you used chatGPT at some point and want to be a kind soul, a (bachelor) student looks into issues of trust, experience and aspects of individual perception and could use your help:
Could you fill out this short (6-7 min) questionaire?
(german or english)
umfrage.hu-berlin.de/17701?src=1DSM
Could you fill out this short (6-7 min) questionaire?
(german or english)
umfrage.hu-berlin.de/17701?src=1DSM
Just to preface my next post, no I did not (just) remember this account to post an advertisement for a questionaire to help a student under my supervision with his thesis...
...but I must admit it was a factor...sorry in advance?
...but I must admit it was a factor...sorry in advance?
February 8, 2024 at 10:31 AM
Just to preface my next post, no I did not (just) remember this account to post an advertisement for a questionaire to help a student under my supervision with his thesis...
...but I must admit it was a factor...sorry in advance?
...but I must admit it was a factor...sorry in advance?
I loved this interview with Cat Bohannon on the Daily Show. She is such a powerfull science communicator.
I did buy the book. Out of interest - not for the other subtle hint...
www.youtube.com/watch?v=5eAa...
I did buy the book. Out of interest - not for the other subtle hint...
www.youtube.com/watch?v=5eAa...
November 8, 2023 at 8:54 AM
I loved this interview with Cat Bohannon on the Daily Show. She is such a powerfull science communicator.
I did buy the book. Out of interest - not for the other subtle hint...
www.youtube.com/watch?v=5eAa...
I did buy the book. Out of interest - not for the other subtle hint...
www.youtube.com/watch?v=5eAa...
I wonder why I never found (or remembered?) this great (and old) definition of AI:
"AI is the study of how to make computers do things at which, at the moment, people are better."
Rich, Elaine: Artificial Intelligence. New York, NY USA: McGraw-Hill. (1983)
"AI is the study of how to make computers do things at which, at the moment, people are better."
Rich, Elaine: Artificial Intelligence. New York, NY USA: McGraw-Hill. (1983)
November 3, 2023 at 5:54 PM
I wonder why I never found (or remembered?) this great (and old) definition of AI:
"AI is the study of how to make computers do things at which, at the moment, people are better."
Rich, Elaine: Artificial Intelligence. New York, NY USA: McGraw-Hill. (1983)
"AI is the study of how to make computers do things at which, at the moment, people are better."
Rich, Elaine: Artificial Intelligence. New York, NY USA: McGraw-Hill. (1983)
I'm hyped to be listening to @emilymbender.bsky.social talking about "Meaning making with artificial interlocutors and risks of language technology" online
at www.hitz.eus/webinar_izen...
The talk is brought to us by HiTZ and features a bonus Cat!
at www.hitz.eus/webinar_izen...
The talk is brought to us by HiTZ and features a bonus Cat!
November 2, 2023 at 3:08 PM
I'm hyped to be listening to @emilymbender.bsky.social talking about "Meaning making with artificial interlocutors and risks of language technology" online
at www.hitz.eus/webinar_izen...
The talk is brought to us by HiTZ and features a bonus Cat!
at www.hitz.eus/webinar_izen...
The talk is brought to us by HiTZ and features a bonus Cat!
Reposted by Jan Kraemer
I like this recent comment from gwern on the Gelman blog:
"For reproducibility, it really needs to be reproducible and checkable by a stranger with little time or energy to spare, because even the author will soon enough be that stranger."
statmodeling.stat.columbia.edu/2023/10/29/s...
"For reproducibility, it really needs to be reproducible and checkable by a stranger with little time or energy to spare, because even the author will soon enough be that stranger."
statmodeling.stat.columbia.edu/2023/10/29/s...
November 1, 2023 at 2:39 PM
I like this recent comment from gwern on the Gelman blog:
"For reproducibility, it really needs to be reproducible and checkable by a stranger with little time or energy to spare, because even the author will soon enough be that stranger."
statmodeling.stat.columbia.edu/2023/10/29/s...
"For reproducibility, it really needs to be reproducible and checkable by a stranger with little time or energy to spare, because even the author will soon enough be that stranger."
statmodeling.stat.columbia.edu/2023/10/29/s...
Reposted by Jan Kraemer
So because I was reminded - the definitive answer to "Is a hot dog a sandwich?" is "by what taxonomy?"
Categories matter, after all. Taxonomies exist for a reason. If you're not clear on defining your categories, then your mixed berry salad will consist of:
Categories matter, after all. Taxonomies exist for a reason. If you're not clear on defining your categories, then your mixed berry salad will consist of:
October 18, 2023 at 5:38 PM
So because I was reminded - the definitive answer to "Is a hot dog a sandwich?" is "by what taxonomy?"
Categories matter, after all. Taxonomies exist for a reason. If you're not clear on defining your categories, then your mixed berry salad will consist of:
Categories matter, after all. Taxonomies exist for a reason. If you're not clear on defining your categories, then your mixed berry salad will consist of:
Habe grade eine Benachrichtigung über eine MeetUp Gruppe in Berlin erhalten, der ich gar garantiert nicht beitreten werde:
"Pro Kapitalisten & konsequente Liberale"
...wobei mich eine anthropologische Studie über die Teilnehmer sehr interessieren würde...
"Pro Kapitalisten & konsequente Liberale"
...wobei mich eine anthropologische Studie über die Teilnehmer sehr interessieren würde...
October 30, 2023 at 12:12 PM
Habe grade eine Benachrichtigung über eine MeetUp Gruppe in Berlin erhalten, der ich gar garantiert nicht beitreten werde:
"Pro Kapitalisten & konsequente Liberale"
...wobei mich eine anthropologische Studie über die Teilnehmer sehr interessieren würde...
"Pro Kapitalisten & konsequente Liberale"
...wobei mich eine anthropologische Studie über die Teilnehmer sehr interessieren würde...
Reposted by Jan Kraemer
This by @jemima.bsky.social is quite good. I know a lot of Gen-Z folks and I can absolutely get their reticence to invest all their time, energy and mental health into a 9-5 corporate job; the people angered by that are really saying the quiet part out loud about what we should expect from work.
In defence of the Gen Z challenge to the ‘work ethic’
[FREE TO READ] Contempt for younger workers protesting at the daily grind masks the fact that so few are happy with it
on.ft.com
October 30, 2023 at 11:37 AM
This by @jemima.bsky.social is quite good. I know a lot of Gen-Z folks and I can absolutely get their reticence to invest all their time, energy and mental health into a 9-5 corporate job; the people angered by that are really saying the quiet part out loud about what we should expect from work.
Reposted by Jan Kraemer
Man fuck this shit. That's my nuanced academic position. Fuck this shit and fuck any medical professional who presents this as a "good."
Hell, I have an entire chapter coming out on why this shit is terrible for disabled people.
Hell, I have an entire chapter coming out on why this shit is terrible for disabled people.
“We want to eventually transition parts of our tech to become stand-alone — to become fully automated and remove the doctor or the nurse from the loop…” 💀
AI has arrived in your doctor’s office. Washington doesn’t know what to do about it.
AI is diagnosing diseases and recommending treatments, but the systems aren’t always regulated like drugs or medical devices.
www.politico.com
October 30, 2023 at 12:27 AM
Man fuck this shit. That's my nuanced academic position. Fuck this shit and fuck any medical professional who presents this as a "good."
Hell, I have an entire chapter coming out on why this shit is terrible for disabled people.
Hell, I have an entire chapter coming out on why this shit is terrible for disabled people.
Reposted by Jan Kraemer
I am always so amazed to learn/realize how many words of yiddish origin are in the german language.
And then I get sad, thinking about the horrors (us) german people inflicted on others. (Others as in non-germans, as well as othered germans, jewish people bein highly targeted in both groups)
1/
And then I get sad, thinking about the horrors (us) german people inflicted on others. (Others as in non-germans, as well as othered germans, jewish people bein highly targeted in both groups)
1/
October 27, 2023 at 8:24 AM
I am always so amazed to learn/realize how many words of yiddish origin are in the german language.
And then I get sad, thinking about the horrors (us) german people inflicted on others. (Others as in non-germans, as well as othered germans, jewish people bein highly targeted in both groups)
1/
And then I get sad, thinking about the horrors (us) german people inflicted on others. (Others as in non-germans, as well as othered germans, jewish people bein highly targeted in both groups)
1/
Reposted by Jan Kraemer
Just remembered that this great post by Richard Morey exists: Why you shouldn’t say “this study is underpowered”
towardsdatascience.com/why-you-shou...
towardsdatascience.com/why-you-shou...
Why you shouldn’t say “this study is underpowered”
Ensuring our critiques of power are relevant, clear, and based on good statistics
towardsdatascience.com
October 25, 2023 at 9:47 PM
Just remembered that this great post by Richard Morey exists: Why you shouldn’t say “this study is underpowered”
towardsdatascience.com/why-you-shou...
towardsdatascience.com/why-you-shou...
Reposted by Jan Kraemer
Also, the psychological burden of carrying those attacks is very real, especially as new graduate. It pushes people out of the field.
In case this helps anyone, as a peer reviewer, I take this approach: 🧵
In case this helps anyone, as a peer reviewer, I take this approach: 🧵
October 26, 2023 at 12:33 AM
Also, the psychological burden of carrying those attacks is very real, especially as new graduate. It pushes people out of the field.
In case this helps anyone, as a peer reviewer, I take this approach: 🧵
In case this helps anyone, as a peer reviewer, I take this approach: 🧵
I wonder how long it will take my surroundings to get used to me saying "There is light on the end of the frog"...
...thanks go to @tkingfisher.bsky.social for making my language weirder and congratulations follow swiftly for her Hugo 🎉
...thanks go to @tkingfisher.bsky.social for making my language weirder and congratulations follow swiftly for her Hugo 🎉
October 23, 2023 at 5:22 PM
I wonder how long it will take my surroundings to get used to me saying "There is light on the end of the frog"...
...thanks go to @tkingfisher.bsky.social for making my language weirder and congratulations follow swiftly for her Hugo 🎉
...thanks go to @tkingfisher.bsky.social for making my language weirder and congratulations follow swiftly for her Hugo 🎉
Reposted by Jan Kraemer
I am finding joy in using generative AI to finally give fictional characters the win they have been desperately seeking
October 23, 2023 at 12:28 PM
I am finding joy in using generative AI to finally give fictional characters the win they have been desperately seeking
If anyone does any work on Folk Theories of Machine Learning, please say Hi and lets maybe chat? (This topic rather central to my current area of interest right now.) 🧪
October 23, 2023 at 12:34 PM
If anyone does any work on Folk Theories of Machine Learning, please say Hi and lets maybe chat? (This topic rather central to my current area of interest right now.) 🧪
Reposted by Jan Kraemer
A somewhat niche blog post on FrontierAI sparked by a comment by Jack Stilgoe blogs.nottingham.ac.uk/makingscienc... #AcademicSky
October 20, 2023 at 6:17 AM
A somewhat niche blog post on FrontierAI sparked by a comment by Jack Stilgoe blogs.nottingham.ac.uk/makingscienc... #AcademicSky
Reposted by Jan Kraemer
Full sequence of the annular solar eclipse at Shiprock, New Mexico. This was a dream composition of mine. So much time, energy, and money went into making it happen. I'm forever grateful to the Navajo for letting me shoot it from sacred ground. An experience I'll never forget.
October 19, 2023 at 2:50 AM
Full sequence of the annular solar eclipse at Shiprock, New Mexico. This was a dream composition of mine. So much time, energy, and money went into making it happen. I'm forever grateful to the Navajo for letting me shoot it from sacred ground. An experience I'll never forget.
Reposted by Jan Kraemer
*sigh*
I am so amazed how weird the apparent standard for publications in the ML/"Deep Learning" field is. I accept that starting from scratch for each paper would be rather tedious (to read and write), but having to parse a stack of 15 arxiv papers just to collect the underlying assumptions?
I am so amazed how weird the apparent standard for publications in the ML/"Deep Learning" field is. I accept that starting from scratch for each paper would be rather tedious (to read and write), but having to parse a stack of 15 arxiv papers just to collect the underlying assumptions?
October 19, 2023 at 9:39 AM
*sigh*
I am so amazed how weird the apparent standard for publications in the ML/"Deep Learning" field is. I accept that starting from scratch for each paper would be rather tedious (to read and write), but having to parse a stack of 15 arxiv papers just to collect the underlying assumptions?
I am so amazed how weird the apparent standard for publications in the ML/"Deep Learning" field is. I accept that starting from scratch for each paper would be rather tedious (to read and write), but having to parse a stack of 15 arxiv papers just to collect the underlying assumptions?
Reposted by Jan Kraemer
I just thought “this sounds like the type of thing I’d make up for a blog post” and lo and behold www.the100.ci/2023/06/13/d...
October 17, 2023 at 6:35 PM
I just thought “this sounds like the type of thing I’d make up for a blog post” and lo and behold www.the100.ci/2023/06/13/d...
Screams to high heaven: please never use healthcare as an area where we could benefit from less regulation wrt. the EU AI act.....misdiagnosis is a problem, but unregulated ai will surely not help. (I had a few too many machine learning in health papers to review to be able to believe that....
October 17, 2023 at 1:02 PM
Screams to high heaven: please never use healthcare as an area where we could benefit from less regulation wrt. the EU AI act.....misdiagnosis is a problem, but unregulated ai will surely not help. (I had a few too many machine learning in health papers to review to be able to believe that....
Trying to send out the #Philosophy bat-signal...
I hope that means Zach is writing ,,AI: Abridged Beyond the Point of usefulness"
🧪 #CogSci
I hope that means Zach is writing ,,AI: Abridged Beyond the Point of usefulness"
🧪 #CogSci
Hey philosophy geeks-- could someone explain Godel's belief that the brain, but not the mind, is reducible to mechanism? I find this sort of dualism just baffling, because surely minds have to operate by some means, even if they're somehow not physically embodied in a skull.
October 17, 2023 at 12:52 PM
Trying to send out the #Philosophy bat-signal...
I hope that means Zach is writing ,,AI: Abridged Beyond the Point of usefulness"
🧪 #CogSci
I hope that means Zach is writing ,,AI: Abridged Beyond the Point of usefulness"
🧪 #CogSci
Reposted by Jan Kraemer
I wrote a thing in response to a thing. blog.liedra.net/i-am-the-enemy
I am the enemy.
OK, so I am the enemy. The enemy of all those who would design, develop, deploy and maintain systems that ride roughshod over people wh...
blog.liedra.net
October 17, 2023 at 11:24 AM
I wrote a thing in response to a thing. blog.liedra.net/i-am-the-enemy