AI Checkers Forcing Kids to Write Like a Robot to Avoid Being Called a Robot

https://www.techdirt.com/2024/09/04/ai-checkers-forcing-kids-to-write-like-a-robot-to-avoid-being-called-a-robot/

by speckx • 2 years ago

88 96 2 years ago

I submitted several scientific/research papers my wife wrote or co-authored in the late 90s early 2000s and they were all flagged as AI written by multiple tools. It was good for a laugh.

fakedang• 2 years ago

That's just more evidence that the Kaylonian robots have already infiltrated our social and familial ranks. /s

oerdier8• 2 years ago

As a teacher, my answer to publicly available AI, regarding grading, has been to only let work by students I have seen been done myself (or staff, trusted implicitly) count towards passing/failing/diplomas.

This means graded longer tasks such as writing or programming, which in the past were done outside of class and then handed in, are instead done in person (like most exams are, traditionally). To acknowledge the longer process, the time alotted for such a test is much longer (up to 3 hours instead of 1) and I can expect less of work created in 3 hours than work created over 6 weeks.

Students still have the opportunity to hand in work that was done outside of such an in-person test, which I will grade, but that grade isn't on the record is serves only as an indication.

I wish I could trust students, but plagiarism has become so easy now (added to a zeitgeist of personal gain above all) that a critical mass fails to resist.

I think diploma's are very useful. They get their usefulness from the trustworthiness of the party handing out the diploma. Only recording grades for work that has been done in an environment controlled by that party has downsides, but in the current and foreseeable situation I don't see a better way of maintaining a base level of trustworthiness.

I love teaching without grades or diplomas involved so much more. Then it's all about doing what's best for the students.

candiddevmike• 2 years ago

Are you seeing a shift in class duration to accomplish this? For my kids, I don't see how their classes can squeeze both instruction time + work time in the allocated slots. Is instruction time being sacrificed or "replacing" homework (watching videos/reading text at home)?

oerdier8• 2 years ago

In principle nothing about the structure of the course has to change. You can keep an existing course with instruction during class and and working on a project outside of class. Even handing in the project and grading could stay the same, although the grade for the project wouldn't be recorded and would serve only as an indication. You then add a longer test at the very end, for a grade.

How I set up my course is different from traditional education. Most instruction happens from self study (including videos and software). Face to face time is for questions. All elements, material, staff, and students, have to be tuned for this setup to work though and is hard to pivot to in existing setups because education (in the Netherlands at least) is very entrenched.

This self-study centered learning approach, like the in-person tests instead of projects, comes from necessity. Great (and I expect ever increasing) teacher shortages.

orwin• 2 years ago

A friend of mine seems to have switched to a "exercise/exam in school, learn at home" methodology, where the next lesson is given in sheet format at the end of the hour, and "school" is mostly exercises he uses to see how much each student understand. It seems to be quite happy with this, and told me good students are quite happy with this (lessons are learned at their own pace), and poor ones now have more time with him helping them during the exercises. Only issue is that he now have troubles not talking to/helping students during exams and now really hate those (always disliked them as well as grading, but now it seems worse).

He also heavily encourage student to learn with each other during "perm" hours (basically in highschool, hours where student have to stay in school but have nothing to do and can work on their own, go to the library or play/chill/whatever), but that was before and because he fondly remember the days where all 26 of us, the full class, all teamed together against homework (yeah, i had great classmates)

JohnFen4• 2 years ago

my answer to publicly available AI, regarding grading, has been to only let work by students I have seen been done myself (or staff, trusted implicitly) count

Of all of the tactics I've heard teachers taking, this is the one that seems both the most effective and most fair. It's a terrible thing that it's necessary because it's a less effective use of classroom time, but it's hard to think of a better approach.

nomel• 2 years ago

My prediction has always been that schools will shift to proctored exams, and job placement rates will eventually be the only metric that matters for students. I think it's an advantage for the schools to ignore AI for as long as they can though, since higher pass rates means more students paying the full cost of tuition, and better graduation statistics.

hobs2• 2 years ago

I think the problem is we think using another KPI will obviously solve the problem. Using job placement as a rate just encourages nepotism and self dealing though - uni's have been caught giving say, an average salary of an outgoing degree by including an NBA player who got it before making millions shooting hoops.

klyrs• 2 years ago

You're basically describing the saga of code bootcamps before they all tanked in a blaze of VC

madrox10• 2 years ago

I think you're absolutely correct.

I worry that the uncomfortable truth AI is revealing is that essay writing is a less essential skill than we'd like to believe. We can say things like "personal gain above all" and the pursuit of a diploma being our downfall, but I can't help but wonder if it would be better if we just got rid of essay writing altogether or wait until the student knows what kind of writing they want to do.

Of course you prefer teaching without grades; the only people in the room are people who want to be there.

ryandrake• 2 years ago

I kind of agree with this and am willing to take it even further: Why should I even study a subject that I can just ask a computer to explain to me when I need it? AI isn't quite there yet in terms of reliability, but there may be a point where it's as reliable as the calculator app, at which point, does it even make sense for me to study a subject just to get to the mastery level that is already matched by an AI system?

If I need to know when Abraham Lincoln was born or when the Berlin Wall fell, I could either 1. memorize it in high school history class to demonstrate some kind of "knowledge" of history, or 2. just look these things up or ask an AI when I need them. If the bar for "mastery" is at the level of what an AI can output, is it really mastery?

jprete• 2 years ago

Really it's not revealing anything at all about the value of essay-writing as a skill. It's just revealing that people will cheat in ways that are hard to directly prove, and grading writing is really hard when people have access to an infinite bullshit generator.

jkestner• 2 years ago

We don't write just to communicate things we know, but to learn what we think. Especially in a school setting.

https://herbertlui.net/dont-think-to-write-write-to-think/

I remember this every time I solve a problem while explaining it to someone in an email.

nerdjon• 2 years ago

I can see the value in this approach, but I also know for myself this would have been horrible in school.

I was always a horrible test taker (the pressure being a large part of that, but not all) and I often needed the time outside of class for any papers I need to write. No matter how much I tried, it just wasn’t happening in class.

I acknowledge the problem but I would be worried how this could negatively impact a lot of people.

oerdier8• 2 years ago

You're right that some students fail only because of the way they are graded. Although with the right preparation this number can be quite low, it's never zero. It's heartbreaking to see some fall between the cracks, but I don't think it's feasible to get to a system that works for every single person.

JieJie• 2 years ago

Right. At my school, the accessibility option for tests is to give us more time. A three hour test would be made "accessible" by extending it to six hours or even more.

I was really hoping AI would make our world more accessible, not less.

(eta) Additionally, it would take more instructor or docent time, because no one can't be trusted to actually learn the material we're paying tens or even hundreds of thousands of dollars for.

It ain't the future dystopia I'm afraid of, it's the one we're creating this week.

rimunroe• 2 years ago

Thanks for bringing this up. I nearly failed out of college before getting accommodations for anxiety. I was a physics major and was solid with math, but sitting down to take a test I’d forget how the most basic properties of multiplication and exponentiation worked. Once I started being able to do untimed take home tests (due to accommodations) I started doing well. With the pressure off I was also able to finish the tests much more quickly than before.

jolmg• 2 years ago

As a teacher ... tasks such as writing or programming

I hadn't realized programming was being taught pre-university. From a quick look online it seems high-schoolers may be learning Python. That's pretty cool. Wonder how widespread this is, and how early children are taught.

quasse• 2 years ago

My high school had several years worth of programming courses available, and that was in the 2000s.

Now days American high schools are becoming much more focused on teaching technical skills again (thankfully). There's a renewed focus on teaching things like CNC machine programming and robotics alongside traditional computer science courses.

ericmcer2• 2 years ago

My kid is in a programming class, so far it has been... kind of a joke. Extremely controlled environments and dragging "code" around in the form of colored blocks.

I wish they taught JS/HTML web dev stuff. Even if someone only gets a year in the knowledge will stay relevant because they are using the internet daily. Just basic understanding of cookies, http, IP addresses, etc. is something kids should know.

oerdier8• 2 years ago

In the Netherlands, programming in secondary education (ages 12-18) is not uncommon but also not widespread. If it's offered, it's not great. Like programming taught at universities, it's mostly people with a knack for it that actually improve.

For my msc thesis, which I finished 6 years ago, I researched how to teach programming to people without natural talent. I found a method with great results, and have yet to see anyone do something remotely similar. I'm not teasing a sale, all my stuff will be fully free. I've been working on it fulltime for the last three years, and I expect to have a 1.0 in two years, so let's say four years.

Terr_• 2 years ago

It's also worth pointing out that this approach is valuable in protecting against some very traditional risks too: Content that was ghostwritten or plagiarized from another human. LLMs simply made it easier and cheaper to cheat in the same sets of circumstances.

but that grade isn't on the record is serves only as an indication.

I think that kind of output is often under-appreciated, especially by students. Even as pure advisory feedback, a grade on a subject or section helps people recognize where to allocate their limited time and effort.

FloorEgg5• 2 years ago

Hey there is a company examind.io that has a product that lets you assign writing assignments for students to do on their own time in their UI that runs in the browser, and it tracks every keystroke and interaction by the student and analyzes it to determine whether it's the student writing or if they are copy/pasting or transcribing it from AI. Also it has option to give student access to AI research assistant right in the software so you can inspect how the student is using AI to help with their work.

I think it might give you the assurance you need while giving the best experience and opportunity for your students to do honest / hight integrity work with the latest (AI) tools.

The benefit of forcing students to write their grades work in person is you know it's coming from them, but the downside is it's a very artificial test not representative or what real world work will be like. I think examind can give you the best of both.

joatmon-snoo• 2 years ago

Assuming you're the founder, this is the type of BS comment that makes the rest of us hate AI founders.

It's vacuous, makes vague claims that don't leave room for proof/disproof, and doesn't offer any reason that it's any better than a prompt that asks GPT4o "was this generated by AI y/n"

em-bee• 2 years ago

i can use AI to generate text and then type it manually into that tool.

oerdier8• 2 years ago

I don't feel assured at all. I don't want to bet on any horse in the volatile arms race between AI and anti-AI.

iambateman• 2 years ago

The question that came to my mind was "how do we give children the opportunity to think through problems which cannot be processed in two hours?"

I don't know the answer.

JohnMakin• 2 years ago

Teach critical thinking before the college level - ideally well before. Much of the issues with overly relying on these tools boils down to a lack of critical thought - analyzing the output of these things critically, once you reach a certain competence of writing, will lead you to quickly realize you shouldn't overly rely on it, or maybe even use it at all.

downut• 2 years ago

I have no idea what it is like now but a significant part of the testing environment at Georgia Tech in the early '80s in the highly competitive engineering fields was "impossible" exams. These involved a timed in-class exam of an hour (sometimes two) where it was not possible to solve the major points problems analytically within the time limit, but a variety of approximation techniques were available. The highest scores came from understanding how to deploy the fastest approximation techniques.

Unlike many of my classmates I was really good at the mathematical part of the exam questions and could straightforwardly solve them analytically. However, this took too much time! I bombed several important exams learning that "good enough" is an important attitude to have when solving engineering problems. I hated this intensely at the time. (Never give that Stalinist institution a dime till I die.) BUT. There's a reason I went into numerical analysis in graduate school. Which I loved then, turned into a transient career that I enjoyed immensely, and remember fondly.

I doubt this approach is feasible for non-engineering disciplines.

[edit: redundancy removal]

more_corn• 2 years ago

Give yourself a few more hours to think about it. These things can’t be rushed.

oerdier8• 2 years ago

It's hard, but you can create tests that you can only pass if before you've processed a problem for weeks.

dianeb• 2 years ago

It's really difficult to plagiarize if you write the paper or take the exam in class using paper and pen-or-pencil. It's also difficult to plagiarize if the teacher/professor/TA actually wanders around the room during the test (without stopping behind a student and raising the student's blood pressure in the process.)

Of course, that involves more on your part as teacher/professor -- you actually have to teach and test and act as editor -- on an ongoing basis. At university, that's part of what TA's are (or should be) expected to do. As a teacher, it's your responsibility.

So, if you happen to be a CS 101 instructor ask for the evolutionary copies of a program. If you happen to teach poetry and cover Herrick, ask for an original poem in the style of "Upon Julia's Dress" and see what the students happen to see from the same poem.

Teaching without ensuring that learning happens is a waste of time and a disservice to your profession.

pants2• 2 years ago

It seems to me that AI detection is best done in the editor itself. If you track a student's mouse and keyboard there should be a clear distinction between someone plagiarizing (even if they're re-typing word for word) vs writing genuine work - which would involve lots of backtracking and editing.

samarthr15• 2 years ago

Oh god, this really does capture my problem. My university demands 40 page reports that are never read by anyone, so my seniors used ChatGPT to generate the needed drivel needed, so now we have the ai detectors, and are told that the % of AI content as flagged by a tool (turnitin) must be below 10%.

Unfortunately, the way I write, (like any well educated Indian) is a rather verbose, formal and rather florid in it's manner, and with no fewer than 3 conjuctives (when I don't want the reader to understand what I just wrote) per sentence. This style is for good or bad understood by turnitin as being AI generated.

So, I have had to unfortunately simplify and write in a more stilted manner, just to be able to submit my assignment.

Frenchgeek• 2 years ago

Train ChatGPT to avoid being flagged by turnitin? (Then write a paper on it in a way designed to trip said AI detector)

samarthr15• 2 years ago

Been there, tried that.

It only sorta worked, but failed basis m basic "formalism" tests. So could not be directly used.

The writing style and vocabulary expected by Indian universities is a style of English that has not been used in England for 80 or so years. It is uniquely frozen in time, florid, and all together horrid. (I had to make it rhyme, sorry)

greiskul• 2 years ago

when I don't want the reader to understand what I just wrote

Unless you are writing poetry or art (where ocasionally this makes sense, to avoid censorship for instance, or to make a point), that's just bad writing. Should you get penalized for plagiarism? No. Should you get bad grades for writing like this? Probably yes.

samarthr15• 2 years ago

It only matters when someone actually reads it no?

If a tree falls and nobody heard it, did it actually fall?

Also relevant, this is for my final year engineering project, not a writing course.

The report that I write is not even read by those who grade my project, rather, the page count is checked, and a small amount of sanity-check-reading to make sure I have not just filled the pages with cinema songs is done. My grade is decided by my relationship with the panel, nothing else, not the quality of work done, not the idea, not the deliverables shown.

I am absolutely sick and tired of this.

I just wanted to do a good job, and write good code, but no, that does not move the needle one bit with our panel of reviewers.

(Pardon my rant, but I really did want to do a great job with my project, but my university did not want to reward it)

At any rate, I can get into a whole rant about incompetent lecturers at many Indian universities, but that is way off topic.

anigbrowl• 2 years ago

There will eventually be a great lawsuit from people have been disenfranchised by these BS 'AI checkers'. Someone will sit down in court and write essays from scratch on a random topic, which will be flagged as AI-generated, leading to general mockery and the diminution of the stock toward zero.

madrox10• 2 years ago

Writing teachers have a lot to learn from Math teachers, who have had to fight against the crutch of calculators for decades. The future here is obvious. More in-class tests to evaluate writing ability instead take home is what will happen. For take home, require "showing your work" and the research that went into writing. Show the steps.

Yes, you could still use AI for any take home work, but I think what forces students to AI isn't the lack of will to do the work. It's the "I have a 5 page paper due tomorrow and I haven't started" cliff.

I mean, writing education is messed up in so many ways, but if I'm being realistic this is the path forward.

ok_dad2• 2 years ago

Seriously, the answer is to have the students turn essays in using stages. First stage is to write on paper by hand, even if there’s not a lot of research or facts yet. Lay out the argument first using what they remember from the initial research, then look it over with the kid and have them explain it a bit. Then, they do the next stage which is to do closer reading and research, and fill in the gaps. Finally, they polish the essay and turn in the final.

It’s hard to use AI to do this, and if a student did use AI effectively like this, they’ll still learn a lot about the topics, because they’re forced to think about it.

Imagine, rather, that you have these kids actively use AI to help them write an essay, but make them provide the transcripts so you can so how they think and use the tool. Perhaps even sit with them and use the AI tools and show how they can be used as a guide, as long as you don’t take the responses as facts like in an encyclopedia, but rather use AI like a parent who is getting but doesn’t quite remember the details. You’d not trust it implicitly but just use as a guide. Perhaps someone should make an AI that isn’t so sure of itself, and suggests ways to research things yourself but points you towards the closest cardinal direction to your question?

saagarjha• 2 years ago

This scares me as a serial procrastinator.

logicchains• 2 years ago

This is an expected consequence of using ML to detect AI. LLMs are more well-read than the average person, so they'll have a bigger vocabulary than the average person, so the ML model will learn to associate a bigger vocabulary with a higher probability of being AI. Which unfairly penalises well-read humans, who also have a bigger vocabulary than average.

Personally I think the whole endeavour is a waste of energy. Unless there's a massive civilisational collapse, humans will always have LLMs as writing aides, so trying to teach them to write without one is teaching them a skill they'll never need to use.

If you want to teach them to reason on the spot, then go back to requiring them to orally put forth and defend an idea.

rootusrootus• 2 years ago

If you want to teach them to reason on the spot, then go back to requiring them to orally put forth and defend an idea.

I have been thinking for the last couple years that we may end up in that spot. It is more labor intensive, and perhaps a little more susceptible to bias, but pretty hard to fake. At the end of the term, each student has half an hour with the professor (or a TA) where they discuss the content of the class. It should be quickly apparent who knows the material well enough and who skipped too many lectures.

actionfromafar• 2 years ago

Ubicomps colonial impulse, back for round 2.

jedberg2• 2 years ago

If you put Shakespeare or the Constitution through one of those, they will say 90%+ AI. Any student who gets dinged by one of those things just needs to show this result until the administration caves and bans the stupid checkers.

They may have to go to the local press or a board meeting to get that result though.

thih9• 2 years ago

Do you have a source on Shakespeare being flagged? I just tested first few lines of Hamlet with a random AI detector[1] and got a result of 100% human.

I saw a post saying that the opposite happens, i.e. AI output, when prompted to write like Shakespeare, can beat AI detectors[2].

[1] https://quillbot.com/ai-content-detector and https://www.zerogpt.com

[2] https://old.reddit.com/r/freelanceWriters/comments/1amngnv/i...

jedberg2• 2 years ago

I don't have a direct link. It came from a reddit post by a professor who was trying to convince his department to stop using them by putting famous works through to show how bad it is.

But my guess is that it works both ways depending on the cheating detector and the prompt used.

mensetmanusman• 2 years ago

So that’s how Shakespeare did it…

ronsor2• 2 years ago

AI checkers need to be sued out of existence.

tombert• 2 years ago

Sued for what exactly? Like what law or regulation are they breaking?

ronsor2• 2 years ago

False advertising and misrepresentation.

Animats• 2 years ago

There is a certain style to LLM-created writing. It tends to float in the air, lightly tethered to real world examples. Unfortunately, that's also what passes for academic style in some quarters. I mentioned that a few days ago here.[1]

LLMs have grammatically correct blithering nailed. One amusing result is that Derrida and his followers now sound like an LLM with the temperature medium high.

Maybe we should be teaching kids to not write like that.

Mandatory XKCD.[2]

[1] https://news.ycombinator.com/item?id=41419645

[2] https://xkcd.com/451/

arjie3• 2 years ago

Interesting that my childhood evaluation systems in India are resistant to this new development. 100% scores from in-person exams, no loss of dynamic range through grading (i.e. 97% score > 96% score), zero score to homework. I imagine that the primary constraint in the US is labour cost since you need what you call a proctor and what we call an invigilator, and perhaps the substantial student power (a student and their parents can overrule teachers).

I like the relative mobility in the Indian system since anyone can swot to the test, but clearly the US system is much more enjoyable for a wider variety of students and they have a wider range of skills.

I think I'd like for my children to have a composite of the two. I dislike neverending homework and love tests because spiky activity allows for lots of self-directed exploration and the tests allow for evaluation. Though it's like that this is because I, personally, am quite capable with tests: I operate very well under time pressure and have an exceptional memory.

samarthr15• 2 years ago

Is the relative mobility a good thing though? (Not strictly on topic, so apologies)

I would argue that for the exams that actually matter (the JEE, say) PCM scores are very poor indicators of success in the chosen branch, and are an arbitrary selection criteria, and not very removed from a lottery.

A better system may include psychometric pairing for branch of engineering opted?

My argument is essentially that if I like (say) Civil engineering more, then I would have demonstrated some additional knowledge in parts of the desired branch, so testing that should in theory give better civil engineering graduates, instead of people who joined it for the tag of the institute and not for the branch?

But again, I am a _bit_ salty about the 12th standard exam treadmill in india.

arjie3• 2 years ago

Yes, it's practically a lottery with a skill floor. And I think that's a pretty good way of doing things. The purpose of these technical institutes is to reliably train people into a bunch of skills, not to get people who have the skills already and stamp them. The purpose of an entrance exam is to have a method that:

- demonstrates a floor of intelligence and skill

- is demonstrably fair

- has high dynamic range

- can be trained against by the conscientious

I think the JEE behaves well against these criteria. In addition, affirmative action is explicitly handled by quotas. This gives a predictable system with known constraints. When participating, you know your targets. Personally, if I know my constraints ahead of time, and I can optimize to them by studying, and the studying is directionally useful to education, that is sufficient to me. If I fail, it's on me and I can accept that.

hobs2• 2 years ago

How can we claim any high ground when we are handing off the need to check for plagiarism to a tool that in effect is the same as the student using it to create plagiarized works.

It doesn't even work, its non deterministic, and most importantly - if the teacher is familiar with the student's writing it should be pretty obvious if this is a fake or not based on something they can be observed doing.

dheera• 2 years ago

OK so back then using SAT words was good, and now using SAT words is bad. Got it.

rafram• 2 years ago

Interesting piece.

I really wonder why these tools even show detection percentages when they’re meaninglessly low. A 17% result from a classifier model is totally meaningless. You could show a well-trained animal classifier the first Google Images result for “cat” and it would still say there’s a 17% chance it’s a cow. Showing low-likelihood predictions will do nothing but freak people out and confuse them.

from-nibly• 2 years ago

Tests should not be a tool for authenticating learning. They were NEVER good tools for this and now its clearer why. Tests should be used for a teacher to help a student figure out where they need to focus.

It should be like getting a blood test from a doctor unless they are testing for drugs why would you try to cheat with someone elses blood?

We don't need to figure out how to sidestep the fundamental flaws with our institutional education we need to find and fix the root cause.

Being "autheticated" as "knowing stuff" (your degree and GPA) doesn't translate into companies being able to tell if you can do a job. It didnt before AI and it doesn't now.

Everyone should stop and think about how dumb it is that children and adults are trying to cheat at learning. You can't cheat learning. So why do people try and succeed? Because we arent actually asking for learning we are asking for something else.

AI Checkers Forcing Kids to Write Like a Robot to Avoid Being Called a Robot

Related Stories