To keep me modest, there had to be a typo in a solemn sentence where I was complaining about the lack of alertness in another human being. Point taken. :-)
Well, given that AI is a very fast calculator, it is being funny accidentally, :)
When I tried out ChatGPT, I was stunned by how glitchy and full of "hallucinations" is was, which is a separate issue from how it is programmed to suit its masters and funders. At the same time, it was being actively used by official entities, or to write legal drafts.... no idea how this can work, I think it can't. It requires double-checking every sentence even if we ignore the intentional "bias"!
Some of the more advanced AI's now seem to be running into a wall. The problem (in a nutshell) is that they depend on recycling their own 'hallucinations' and the outcome is often hilarious.
For example; when endlessly generating pictures of elephants they all start to look the same in size and color - however now all elephants now have five legs and 11 trunks.
This makes me wonder if there is limit on how far intelligence (any intelligence) can advance. Even at the best of times our minds merely hallucinate that we are normal.
did you hear about the lawyer who took his unwinnable case (because of statute limit exhaustion) to court? he was disbarred because his AI created supporting law and precedences. (true story.)
Wow. No, I didn't hear, makes sense! I wrote an article about AI in law last year or so, and I was amazed to learn how widely AI is allegedly used. It didn't make sense to me precisely due to "hallucinations" (when the robot just makes stuff up and mixes it with actual facts or at least conventional facts).
yes. jeff childers shared that and found it most puzzling that the lawyer didn't actually check the law cited.
clif high gave an excellent lesson on ai 'wokery' when he repeatedly corrected its 'assumptions' about the relative costs of electric vehicles versus petroleum ones. it was pretty amazing to go from saving something like 10k to costing 100k more. (i forget the actual numbers.) it was fascinating to watch the ai thank clif for each correction and then update the costs.
i shared that with a local musk-ophile and he wouldn't even consider it and dismissed it as some angry republican trying to smear musk's good works. i also found that behaviour fascinating because the 'story' was much much bigger than that and he effectively removed from possible conversation something interesting. he is mostly awake although reluctantly catching up. in 2022 he dismissed 9/11 conspiracy. then he texted me 'you know about tower 7, don't you?' 'yes.'
I hope your friend's journey puts in more in touch with facts, and I hope it is good for your friendship also! :)
As far as ChatGPT, the short amount of time I spent playing with it was a disappointment. It couldn't help lying, and by that I mean pointless lying, not even bias, just making things up. But in order to know it was lying, one would actually have to know the subject matter, and that would defy the purpose of using it for "review of available data." It was more like a fancy word generator, end of story. Which is what it is, at the end of the day.
my thoughts of it looking from outside, so far. i have thought of using it to create an email compatible list of names for a mail out - such as 'go get all the names and email addresses of the mps/mlas in canada in a format compatible for gmail mail out.' i did that manually in 2021 and it was quite a lot of tedious work, of course. (and after that left canada as a refugee!)
as to the friendship, he is struggling with vegan ideology, a much bigger and more difficult, perhaps 'intractable', problem. despite serious health issues, he's doubling down on his vegan practices as if the last eight years of diligent veganism hasn't been enough - 'if only i do what i've been doing harder/purer/stricter'. hmmmm.
so, for me he has given me some really great awarenesses especially around money — he has an amazing history around that, getting clinton to sign a bill that stopped logging in america's public forests — and also an opportunity to practice being open with his journey without the need for me to push him along faster than he goes
as to musk he says 'i know, i know, what is it about that mind-link and experimenting on animals?' and for now keeps the two 'things' of musk split - an odd kind of projection of his own split around that, i suspect. all the best.
When a driver trusts the machine too much, bad things can happen.
Fortunately it doesn't matter when an academic paper trusts the machine too much, and these "researchers" know how the game works.
95% of published papers are just tokens inserted into the tenure slot machine. After you insert 100 tokens, you pull the lever and get a promotion or a degree. Nobody reads what's on the tokens, just as nobody examines the pictures on dollar bills.
I agree!! I have a feeling that as people get sicker and less trained to think from the inside, there will be more of this kind of thing in the physical world. And it is not great!
Just like no one looked at the mortgages contained in the mortgage backed securities that caused the 2008 melt-down.
OK, not /exactly/ like that case. In the MBS case no one even wanted to know what they were already pretty sure of. Well... Maybe it's not really different, come to think of it.
Wow. Astonishing that not only did this get through peer review -- thus demonstrating that nobody really *does* read these papers start to finish -- but it is still up on the website, unedited, with no mention of the inclusion of such an embarrassing oopsie. At least the rat penis article was redacted:
(Warning: the above article contains images of unintentional hilarity.)
I wish I could say that this recent rash of AI-generated nonsense will finally topple the edifice of peer review once and for all, but ... Sadly, I know all too well how entrenched this process truly is in the minds of the so-called "experts." Now, the real question is: will this be one more nail in the coffin of academia, helping the average citizen to wake up to the falsity of the "expert" concept in the first place?
I guess this was an editing error based on how the previous sentence just cuts off and starts the highlighted section. Maybe they were using assistance from AI to write the article and did a poor job editing. However, there is an intriguing allusion that may or may not be unintentional; Iatrogenic cause==human error therefore AI to the rescue.
That's the "funny" part!! Iatrogenic death is the #1 killer, if you consider that CVD & cancer are due to mismanaged health (ok, add a bit glyphosate here & there, & a bit of medical quackery.) The fact that a lot of research is bunk doesn't help matters.
I agree that Iatrogenic death is the number one killer. I expect that the term Iatrogic had been pretty effectively supressed from dialog via corrupt journal publications and now will become more commonly seen. I think it must be very difficult to publish a scientific article with that word in the title. I would expect that word to be filtered out as part of the editing. I would also expect the editor to catch the essentially incomplete paper. It kinda reads a bit like an advertisement for the necessity of infant invasive surgery and requisit medical products. Gross.
Tessa: I knew before the clown show that both scientific and medical publishing were irreparably broken, but now they're not even trying to hide it. "Artificial Intelligence" has to be one of the most ridiculous oxymorons ever invented. How many people actually take this silliness seriously? I'm afraid too many, but the good news is that the number of the red-pilled has grown remarkably large. The panic the elites are in now far exceeds their panic in 2016. It's going to get ugly, but we'll win this one.
this had me laughing. inappropriately again. and took me back to the days when i studied latin, and a part of that was to be introduced to what was, back in the day, the bar jokes: a bear and a greek were at the bar and... well, i don't remember the punch line and no longer have my latin text.
so, i went looking for it and it turns out that creatures of all sorts have been going into bars for ... well likely before written text. here is a look at what is being claimed as maybe the oldest bar joke in written history. i found the discussion around it and meaning to be fascinating; the 'challenge' of context, and just how that core human 'thing' beneath all our bs is there. amazing.
To keep me modest, there had to be a typo in a solemn sentence where I was complaining about the lack of alertness in another human being. Point taken. :-)
lol! yes. ah, life is sweet, with such a great sense of humour.
Indeed! :-)
I used to peer review papers until I realized that it was a waste of my time. AI is obviously much better suited to perform that task.
What I like about AI at this stage is that it is being funny without realizing it. Can't wait to see what happens if AI masters sarcasm.
Well, given that AI is a very fast calculator, it is being funny accidentally, :)
When I tried out ChatGPT, I was stunned by how glitchy and full of "hallucinations" is was, which is a separate issue from how it is programmed to suit its masters and funders. At the same time, it was being actively used by official entities, or to write legal drafts.... no idea how this can work, I think it can't. It requires double-checking every sentence even if we ignore the intentional "bias"!
Some of the more advanced AI's now seem to be running into a wall. The problem (in a nutshell) is that they depend on recycling their own 'hallucinations' and the outcome is often hilarious.
For example; when endlessly generating pictures of elephants they all start to look the same in size and color - however now all elephants now have five legs and 11 trunks.
This makes me wonder if there is limit on how far intelligence (any intelligence) can advance. Even at the best of times our minds merely hallucinate that we are normal.
I prefer human interaction, as beautifully imperfect as it is
Somehow I think of Dr Spock and sarcasm when I read this.
Dr. Spock or Mr. Spock?
Haha right. I was too young to be a Trekkie.
did you hear about the lawyer who took his unwinnable case (because of statute limit exhaustion) to court? he was disbarred because his AI created supporting law and precedences. (true story.)
Wow. No, I didn't hear, makes sense! I wrote an article about AI in law last year or so, and I was amazed to learn how widely AI is allegedly used. It didn't make sense to me precisely due to "hallucinations" (when the robot just makes stuff up and mixes it with actual facts or at least conventional facts).
yes. jeff childers shared that and found it most puzzling that the lawyer didn't actually check the law cited.
clif high gave an excellent lesson on ai 'wokery' when he repeatedly corrected its 'assumptions' about the relative costs of electric vehicles versus petroleum ones. it was pretty amazing to go from saving something like 10k to costing 100k more. (i forget the actual numbers.) it was fascinating to watch the ai thank clif for each correction and then update the costs.
i shared that with a local musk-ophile and he wouldn't even consider it and dismissed it as some angry republican trying to smear musk's good works. i also found that behaviour fascinating because the 'story' was much much bigger than that and he effectively removed from possible conversation something interesting. he is mostly awake although reluctantly catching up. in 2022 he dismissed 9/11 conspiracy. then he texted me 'you know about tower 7, don't you?' 'yes.'
i haven't played with ai yet myself.
I hope your friend's journey puts in more in touch with facts, and I hope it is good for your friendship also! :)
As far as ChatGPT, the short amount of time I spent playing with it was a disappointment. It couldn't help lying, and by that I mean pointless lying, not even bias, just making things up. But in order to know it was lying, one would actually have to know the subject matter, and that would defy the purpose of using it for "review of available data." It was more like a fancy word generator, end of story. Which is what it is, at the end of the day.
yes.
my thoughts of it looking from outside, so far. i have thought of using it to create an email compatible list of names for a mail out - such as 'go get all the names and email addresses of the mps/mlas in canada in a format compatible for gmail mail out.' i did that manually in 2021 and it was quite a lot of tedious work, of course. (and after that left canada as a refugee!)
as to the friendship, he is struggling with vegan ideology, a much bigger and more difficult, perhaps 'intractable', problem. despite serious health issues, he's doubling down on his vegan practices as if the last eight years of diligent veganism hasn't been enough - 'if only i do what i've been doing harder/purer/stricter'. hmmmm.
so, for me he has given me some really great awarenesses especially around money — he has an amazing history around that, getting clinton to sign a bill that stopped logging in america's public forests — and also an opportunity to practice being open with his journey without the need for me to push him along faster than he goes
as to musk he says 'i know, i know, what is it about that mind-link and experimenting on animals?' and for now keeps the two 'things' of musk split - an odd kind of projection of his own split around that, i suspect. all the best.
When a driver trusts the machine too much, bad things can happen.
Fortunately it doesn't matter when an academic paper trusts the machine too much, and these "researchers" know how the game works.
95% of published papers are just tokens inserted into the tenure slot machine. After you insert 100 tokens, you pull the lever and get a promotion or a degree. Nobody reads what's on the tokens, just as nobody examines the pictures on dollar bills.
I agree!! I have a feeling that as people get sicker and less trained to think from the inside, there will be more of this kind of thing in the physical world. And it is not great!
Just like no one looked at the mortgages contained in the mortgage backed securities that caused the 2008 melt-down.
OK, not /exactly/ like that case. In the MBS case no one even wanted to know what they were already pretty sure of. Well... Maybe it's not really different, come to think of it.
Paper is here. Look for the paragraph near the end starting with “in summary”: https://www.sciencedirect.com/science/article/pii/S1930043324001298
Yep, thank you, I am linking to it in the article as well!
Lazy asses.
Wow. Astonishing that not only did this get through peer review -- thus demonstrating that nobody really *does* read these papers start to finish -- but it is still up on the website, unedited, with no mention of the inclusion of such an embarrassing oopsie. At least the rat penis article was redacted:
https://mashable.com/article/ai-rat-penis-diagram-midjourney-science
(Warning: the above article contains images of unintentional hilarity.)
I wish I could say that this recent rash of AI-generated nonsense will finally topple the edifice of peer review once and for all, but ... Sadly, I know all too well how entrenched this process truly is in the minds of the so-called "experts." Now, the real question is: will this be one more nail in the coffin of academia, helping the average citizen to wake up to the falsity of the "expert" concept in the first place?
It does demonstrate that!! And OMG I am nearly rolling on the floor, the image is epic!!
How hilarious! NOT! What was the doctor doing causing an injury to the liver of a newborn in the first place?
That part is definitely not even remotely funny, I agree with you on that!
lol see also https://mathewaldred.substack.com/p/who-is-editing-these-journals
Also Doctor Kitt just posted this comment, what's inside is too funny
https://tessa.substack.com/p/team-of-researchers-and-ai-walk-into-a-bar/comment/52861115
thanks lmao
blast from the past https://physics.nyu.edu/sokal/weinberg.html
Thank you!
I guess this was an editing error based on how the previous sentence just cuts off and starts the highlighted section. Maybe they were using assistance from AI to write the article and did a poor job editing. However, there is an intriguing allusion that may or may not be unintentional; Iatrogenic cause==human error therefore AI to the rescue.
That's the "funny" part!! Iatrogenic death is the #1 killer, if you consider that CVD & cancer are due to mismanaged health (ok, add a bit glyphosate here & there, & a bit of medical quackery.) The fact that a lot of research is bunk doesn't help matters.
I agree that Iatrogenic death is the number one killer. I expect that the term Iatrogic had been pretty effectively supressed from dialog via corrupt journal publications and now will become more commonly seen. I think it must be very difficult to publish a scientific article with that word in the title. I would expect that word to be filtered out as part of the editing. I would also expect the editor to catch the essentially incomplete paper. It kinda reads a bit like an advertisement for the necessity of infant invasive surgery and requisit medical products. Gross.
Just when we thought things couldn't become more bizarre. On the other hand, these folks are actually in charge. Yikes!
I hear you, Gary!
Doesn't get more embarrassing. Fuck
Chat GPT overuses the word ‘delve.’ https://twitter.com/jeremynguyenphd/status/1774021645709295840?s=12
That is an interesting observation (giveaway?)
Tessa: I knew before the clown show that both scientific and medical publishing were irreparably broken, but now they're not even trying to hide it. "Artificial Intelligence" has to be one of the most ridiculous oxymorons ever invented. How many people actually take this silliness seriously? I'm afraid too many, but the good news is that the number of the red-pilled has grown remarkably large. The panic the elites are in now far exceeds their panic in 2016. It's going to get ugly, but we'll win this one.
this had me laughing. inappropriately again. and took me back to the days when i studied latin, and a part of that was to be introduced to what was, back in the day, the bar jokes: a bear and a greek were at the bar and... well, i don't remember the punch line and no longer have my latin text.
so, i went looking for it and it turns out that creatures of all sorts have been going into bars for ... well likely before written text. here is a look at what is being claimed as maybe the oldest bar joke in written history. i found the discussion around it and meaning to be fascinating; the 'challenge' of context, and just how that core human 'thing' beneath all our bs is there. amazing.
https://www.reddit.com/r/AskHistorians/comments/tbgetc/this_bar_joke_from_ancient_sumer_has_been_making/
Wow!
Perhaps a glimpse into our future.
Or our present. :)