I doubt that anyone is surprised that generative AI is used all over for cheating purposes, right? High school book reports, research papers etc…such that thete are now programmes available to scan documents for plagiarism. I didn’t know this but husband does…luckily. We’ve had an example to handle this past week.
We’re in the final stages of the preparation for daughter’s divorce with the final hearing due to begin 3rd December. Reports on finances, fitness to parent etc have come in…and are surprisingly favourable to the toss pot. Long story not quite so long, we got a chance to eyeball the psych reports. Daughter and husband scan read and skipped to read final recommendations and were poleaxed. Me…once I’d picked myself off the floor…set to scrutinizing each document closely. All their psych evaluations began with one or other questionnaire that were basically the personality tests that I remember reading about folk taking as part of job interviews etc…maybe 1980s?..but are now administered online beforehand (like say a fuller version of the stuff a new PCP sends you) I basically didn’t even recognise the upstanding bloke who’s been abused for most of the marriage by my narcissistic daughter…you can see where I’m going.
I read through his answers for one quiz and, boy, did he sound erudite in his description of narcissistic personality disorder and coercive control, gaslighting and all manner of terminology I’d have to look in a textbook or online source to define. That’s when I smelled fish. Showed it to my husband and he found it equally implausible, ran it through his Grammarly programme (or something similar) and, lo…a 90%+ plagiarism content!!! Lawyer and team informed and we’re starting to attempt damage control today.
How come this psych evaluator didn’t spot it. One of his complaints during interview was that my narcissistic Harvard grad of an abusive daughter routinely denigrated his lesser education…pouring scorn on his speech and writing…and here he’s composed these learned, insight based answers that’d score top marks in a psych exam when in the privacy of his own home without oversight.
I suspect this will become an arms race in the near future, like virus and anti-virus software is today. There will be ever-improving AI and the need for ever-improving AI detection.
Yes, AI can be misused in a variety of situations.
Given a list of key words you are not surprised to see AI use them to create a draft document for proof reading. If done cleverly you might not notice. If the AI part becomes obvious it has gone too far. Ie pages of examples of misbehavior when a few sentences or paragraphs would be typical of human response.
Where do you draw the line? When is AI assistance allowed? And when is it excessive?
Assistance improves productivity but it can be over done.
Nothing new under the sun. I’ve had friends go through messy divorces where children were involved and a similar scenario played out. AI is not doing anything new, just making it faster and easier to do.
Yes. This misuse is actually within an assessment for domestic violence (and yes…I’ve found out belatedly there was)…and a big component in determining custody arrangements. It skewed the relative “truthfulness” score for both of them to the extent that the evaluator took it upon herself to scrap pretty much the whole of my daughter’s testimony (yep…she’s on my “to deal with” list)
Thing is, it’s so blatant with long passages of verbosity…including the non-gender specific pronouns such as they, them, their, and the kids when referring to my granddaughter. Srsly, who talks like that. The evaluator who bolloxed this assessment up needs some serious remedial work.
Don’t even start me on his 'roid use in what’s probably undiagnosed and inadequately treated bi polar II. Husband’s going after that too!
We’re off to meet with her “divorce doula” to put some meet on the arguments for the lawyer to use. This would be interesting…if it were happening to someone else.
Reading just these few comments and listening to husband and the problems with journal editorship, along with recalling all those with star billing on Retraction Watch because of plagiarism gives me a bit of hope that this isn’t quite the isolated example in the family court that I first imagined.
Mini update: …it’s as if no-one had ever heard, let alone entertained the idea of GenAI for gaming these questionnaires. Or even wondering how come this guy could craft such elegant and descriptive prose, when one of his abuse claims was that his Harvard educated wife routinely rubbished his poor writing style.
Short term…all evaluations have been thrown out. Parental Responsibility Evaluation psychologist (child custody) has accepted accountability for the referral to this Domestic Violence Evaluator (who she confidently expected to give her what the court would accept and recommend full custody rather than 50:50) DVE is , in turn, apparently in CYA mode.
To think, I imagined I was wasting my time. Irritating nitpicking…a.k.a Scientific Evaluation…for the win.!
Congratulations Mama Bear. I’m sure you are looking forward to a simpler time in your daughter’s life where she doesn’t need you so much, but strong applause for being there when needed.
IP,
who finds it humorous when parents think their job is done when the kid turns 18
Thanks. Kudos to the husband who decided to run Grammarly and found the plagiarism. I’ve now been tasked with finding examples of why these answers the Lump of Foul Deformity gave should’ve been an immediate Red Flag without resorting to fancy computer analysis…just in case the judge is a dunderhead on these matters.
Honestly, the average 6th grade teacher with no computer savvy would read through these examples of professional level prose…with bullet points listed under headings that might be listed on Wikipedia, say, when discussing Stark’s Coercive Control hypothesis … and smell fish (DVE chick mentioned the use of that on the evaluation…I’m wondering if there was any written background on the test about to be undertaken when they received it)
Given that the primary goal should be to get your GrandD placed properly, how does this aid that? Sounds as though you are going after the person doing the recommendation, rather than focusing on the EX. Am I missing something? Probably. Have been following but not with intense scrutiny. Sorry you guys are going through all this, but would encourage you to stay focused on the primary goal.
That’s a good question…my husband certainly thinks so as he’s been asking it all morning, since the offending reports have been removed as evidence.
It’s a sort of belt and suspenders approach on daughter’s legal team’s part to make sure it’s also cleared from the judge’s mind. The reports were all due in last week and this has given the judge plenty of time to read and digest the damning statements and conclude the wrong conclusion before final hearing even begins. In temporary orders hearings so far, she’s shown herself to be a “both parties are equally bad players” sort of decision maker when making stipulations that don’t require the sort of evidence needed next week. I suspect the team is looking to show that, far from needing what might be construed as futuristic technology to diagnose a pathologic liar and cheat, just simple common sense does a Big Reveal…let alone the clinical acumen of a psychologist posing as a domestic abuse counselor.
Just chiming in here as someone who is involved in AI, the generated AI texts that you and even specialized AI detection applications can not detect and which is what is already out there is what you should really be worried about. Trust me when I say that in a couple years time, you will no longer have any idea if what is presented to you through any form of media is real or not. There are tens of thousands of AI “influencers” posing as people on instagram with millions of followers raking in big bucks from sponsors… And the power of AI is growing more than exponentially where soon they will know us better than our selves as (especially the younger generation who happily hand over their their private thoughts, personal chats, their wants and needs to AI platforms like instagram, facebook, snapchat, chatbots, “free” AI -driven apps on their phones. I wish there was more education about this but technology is always ahead of legislation and we are moving at lightspeed.
Sorry if I go of on a bit of a rant here but if you really want to know what the scale of the impact is going to be like I suggest you watch this presentation from Mo Gahwdat, ex chief at Google and author of One Billion Happy. He was right at the birth of AI and is now travelling the world in the hopes to change people how they use AI. Because it’s not AI that’s inherently bad, it’s people who use it in bad ways which can actually corrupt AI systems in the very general sense.
Here’s the video, he is not promoting a doom scenario but he is simply educating on what is already happening in the AI space right now and postulating a future where AI is omnipresent, which is inevitable.
Sorry if this is off topic but I feel a need to educate people on the topic as this is not common knowledge especially if you don’t work in IT. And to be clear I am not a consipracy theorist
Watch only if the topic interests you, but he explains in a very accessible way what is AI, how it evolves/evolved and where it is probably heading.
Oh one final link here, write an AI text using chatgpt or claude or another of your favorite AI writers/chatbots and run it through a service like https://undetectable.ai/ (click humanize). This is litterally AI combatting AI combatting AI.
Funny you should chime in like this, NoFoolBowser as, when we first fathomed how the albatross had pulled this stunt and expected all and sundry to dismiss the idea, I happened to say that, if this is the first time this judge has come across this sort of thing, she’d better take note as in a year or so, she’ll have more experience than she cares for.
Talking about the judge…maybe she is a bit less “middle ground” than I thought. At an emergency zoom meeting this morning, the Canker Sore’s team presented a rebuttal to my daughter’s forensic accountant’s business evaluation. FWIW, a sum total of $600k due to the knock on effects of systematic embezzlement and theft. Their expert witness countered with just over $6mill. Still a paltry figure compared to what it should be worth. Judge threw it out. Spotted straight away that it wasn’t a rebuttal per se…which, by definition, should be a counter to the figures presented, right?..but rather a non-too-subtle attempt to slip a set of new, dodgy figures and creative accounting practices into the mix. Obviously the expert witness is a bloke after the Toss Pot’s heart!
Not breathing easy yet. Plenty of time for them…Mr GenAI and his team of pettyfogging shysters…to cause more mayhem.
Well, he wrote in his blurb to this DV evaluator “I’ve recently become aware that I’ve been manipulated throughout my marriage by a toxic narcissist”. He said something similar in the pre-trial deposition. When asked about the discovery by daughter’s lawyer, he replied that he’d researched it.
I remember that I’d seen a book lying about on Narcissistic Personality Disorder a few weeks before Thanksgiving last year (and thought he’d been trying a spot of self diagnosis) Didn’t spot any issues of Cosmo so maybe he jumped straight into the research at PhD level with The Book.
Can’t make this stuff up. There was another book lying around on polyamory (another popular Cosmo topic, I’ll wager)…that seems to be all the rage. Well, making out it’s something new is.
A friend of mine took up a lawsuit against a powerfulmMulti-national on the account of #metoo related behaviour by her only superiour, she was pretty much head of Cybersecurity operations Europe. She explained the board and all c-level is typical old boys network, no women, no people of color.
She hired a lawyer who was willing to take on the case but clearly wasn’t competent enough in the end for such a complex case, but my friend was under huge time pressure because she was one month away of becoming a permanent resident here and they wanted to fire her right before that. She started using ChatGPT to do research and find holes in the defense of the firm’s but also on what areas she had a strong case and how to present it. It basically helped her do the lawyers work (who charged an enormous hourly rate of course) for a large part and it helped her win the case. So yeah, it can be used for good. The question is, how do you keep it contained so that countries who don’t take civil rights so seriously for example and don’t place ethical restrictions on their AI systems, scariest being military use. Hence the huge chip ban to China for example. But it would be ignorant to believe the US military isn’t one of the largest investors in AI globally.
Sorry it feels like I am highjacking this thread, still need to get used to the way this forum works. I though I replied to another topic to be frank so apologies.
