UCT scraps AI cheating detection

[Kosmik]

Interesting to see that folks believe tertiary education is about parrot/learning vs application and innovation.

Maybe it's just the field but one of the scariest things we used to hear from engineering was..."open book test".

 

[cguy]

Interesting to see that folks believe tertiary education is about parrot/learning vs application and innovation.

Maybe it's just the field but one of the scariest things we used to hear from engineering was..."open book test".

We had several open book CS exams back in the day. I wonder if they still do that.

 

[Hamster]

Oh lovely. Vibe coders, doctors and engineers incoming...
Patient will be bleeding out and the doctor can't do anything because the "system is down"...

"You are an experienced ER doctor who has watched all the seasons of both House and Greys Anatomy. You are capable of keeping calm under high pressure situations and deliver the best care for any patient in the ER which will result in the highest probability of the patient surviving with as many of their original limbs and organs still attached to their body and functioning optimally.

You are presented with a patient bleeding out of their neck because of a suspected kni...

The patient has lost consciousness. There IS FOAM... OMG there is foam coming out of their mouth. PLEASE ASSIST RIGHT NOW!"

Great question! You are asking for assistance in saving a patient's life. I've prepared a three page essay on possible solutions, let's dig in!

"No! Too much wrods! Please provide steps in concise bullet point format"

Certainly!

- find the wound
...thinking

 

[aleksandar]

My eastern European study experience:

Various project were precursor so you can write exam then after you pass written exam you have oral exam.
Professor, you and board.
Hated few of them because they would look at your written exam and based on that give you questions, stuff that you did not do well. Mathematics prof was known for that.

Personally worse one was operating systems exam (he was at the time also lecturing at UCLA) where during oral exam and discussion prof started asking some question that were related to computer architecture.
He failed me saying that I should known that better and that he will next time ask me that first.

Now architecture prof was interesting because he wold ask you first questions for pass, then he would offer questions for higher grade. You could bail out and get pass or go higher up.
Caveat was that if you did not answer those questions correctly he would fail you.

Definitely lost few years of my life due to stress.

 

[Swa]

If your issue isn't stock standard, most doctors have never, and would never be able to spend as much time on your (or your partner, or kids') particular issue(s) as you can. Between forums, online research, and LLM's, I think it is very reasonable that once your doctor/specialist has hit a wall, that you (as someone capable of understanding and contextualizing the data sources) are more likely to find a root cause or treatment by your own agency, then by continuously begging for tidbits of hope, asking them "What next?", after each failure.

I have had several successes at this myself for sick family members.

In my experience the doctors eventually get tired of you saying that their tests don't show anything if you leave it up to them.

 

[Forum Reader]

In what courses do you have to write essays? English? Even if Chat GPT was around back in my Uni days, I wouldn't have been able to use it for anything. Everything was calculations.

 

[Mirai]

My wife was only diagnosed with Endrometrias, after I read a thread on it in these forums. The symptoms matched, we went to a professor who specialised in it and it was confirmed, after an operation.

Previously doctors just basically downplayed her concerns.

Endometriosis.

 

[Fulcrum29]

We had several open book CS exams back in the day. I wonder if they still do that.

The last time I had an open book test was at school, but the purpose was application. In that context, "open book" is a technical resource.

 

[cguy]

In my experience the doctors eventually get tired of you saying that their tests don't show anything if you leave it up to them.

Yup, and then they start prescribing either painkillers or antidepressants to try get you off their back. They also use terms like idiopathic and neurological, as catchalls when they don't know the cause and/or nature of the issue. I've seen this done for someone with gall bladder dysfunction (needed to be removed ASAP), and someone who had undiagnosed colon cancer.

 

[cguy]

In what courses do you have to write essays? English? Even if Chat GPT was around back in my Uni days, I wouldn't have been able to use it for anything. Everything was calculations.

An essay may be on an HCI analysis of some device or interface, describing the difference between different programming paradigms, describing OSI stack, how the IP protocol works, etc.

Also, LLMs are pretty good at things way beyond essays today. They can write code, and under the correct guidance, do calculations, or output code that would do the calculations.

 

[Swa]

The last time I had an open book test was at school, but the purpose was application. In that context, "open book" is a technical resource.

Only open book test I had was for Computer Olympiad but we were only allowed to use the classroom references as no program examples were allowed. Ended up not using any of them.

 

[Mariner]

Most (not all) degrees aren't worth the paper they're written on these days as information is so easily and readily available.

Sure, all the information is easily available. You can read every available bit of medical or engineering information. It doesn't mean that you will be allowed to practice as a medical professional or engineer though. Just like with a driver's licence that you can read everything about, your competency to drive has not yet been proven.

Would you be comfortable with a neurosurgeon operating on you while Googling or using AI to decide what to do next?

 

[Mariner]

If your issue isn't stock standard, most doctors have never, and would never be able to spend as much time on your (or your partner, or kids') particular issue(s) as you can. Between forums, online research, and LLM's, I think it is very reasonable that once your doctor/specialist has hit a wall, that you (as someone capable of understanding and contextualizing the data sources) are more likely to find a root cause or treatment by your own agency, then by continuously begging for tidbits of hope, asking them "What next?", after each failure.

I have had several successes at this myself for sick family members.

Exactly this. If you're living with an ailment, you're experiencing it often. A doctor has the space of an appointment to try to diagnose the ailment and the patient will not tell him all pertinent information, as it only strikes him when he is experiencing it.

 

[Segg]

Sure, all the information is easily available. You can read every available bit of medical or engineering information. It doesn't mean that you will be allowed to practice as a medical professional or engineer though. Just like with a driver's licence that you can read everything about, your competency to drive has not yet been proven.

Would you be comfortable with a neurosurgeon operating on you while Googling or using AI to decide what to do next?

I'd hope I have a doctor that's not only book smart, but also practically smart and hopefully has a few years of hands on experience

Like I said, not all degrees, but I do have a general distaste for any "business" or "arts" degree, as they're too generic.

 

[geezer]

I had the same GP for more than 40 years. The doctors' offices where he practiced was sold and then he decided to retire, I just stuck to the same doctors' offices and went with the newly appointed doctor. I also had to inform my medical aid of my new preferred general practitioner (PGP). That was in the beginning of this year. At around February I had to go to this new doctor to have my chronic medication script renewed. This new doctor had really weird accent, definitely not from South Africa. He literally sat there and Googled al my symptoms there right in front of me, and also when I told him what medication I am normally prescribed, he also Googled that. He just sat there, listened and Googled after each question. He did not check my BP or listened to my heart or chest. He wrote me a repeat script without even sending me for blood tests.

Unfortunately, according to the medical aid rules, I could only change my PGP after 6 months, which I have done before I needed a new repeat script for my chronic medication.

I had some serious concerns about his medical qualifications. I heard many people refer to him as Dr. Google and the go-to guy if you needed a script or a sick letter for your employer without too many questions asked.

 

[Mirai]

I had the same GP for more than 40 years. The doctors' offices where he practiced was sold and then he decided to retire, I just stuck to the same doctors' offices and went with the newly appointed doctor. I also had to inform my medical aid of my new preferred general practitioner (PGP). That was in the beginning of this year. At around February I had to go to this new doctor to have my chronic medication script renewed. This new doctor had really weird accent, definitely not from South Africa. He literally sat there and Googled al my symptoms there right in front of me, and also when I told him what medication I am normally prescribed, he also Googled that. He just sat there, listened and Googled after each question. He did not check my BP or listened to my heart or chest. He wrote me a repeat script without even sending me for blood tests.

Unfortunately, according to the medical aid rules, I could only change my PGP after 6 months, which I have done before I needed a new repeat script for my chronic medication.

I had some serious concerns about his medical qualifications. I heard many people refer to him as Dr. Google and the go-to guy if you needed a script or a sick letter for your employer without too many questions asked.

That's scary if true. Here's a link:

https://www.hpcsa.co.za/complaints-investigation

 

[Mirai]

Yup, and then they start prescribing either painkillers or antidepressants to try get you off their back. They also use terms like idiopathic and neurological, as catchalls when they don't know the cause and/or nature of the issue. I've seen this done for someone with gall bladder dysfunction (needed to be removed ASAP), and someone who had undiagnosed colon cancer.

Idiopathic is a term to describe conditions for which the causality is either incompletely or totally not understood by medical science in general.
It's not usually used in cases where the doctor can't figure out what's wrong with you but I suppose sometimes people use it inappropriately.

Neurological means associated with the nervous system. Maybe you mean psychosomatic?

The problem with many conditions is that the symptoms are often vague and could imply something serious or could be innocent.
Tests also can provide vague results.
And unnecessary blind testing is not encouraged.

 

[cguy]

Idiopathic is a term to describe conditions for which the causality is either incompletely or totally not understood by medical science in general.
It's not usually used in cases where the doctor can't figure out what's wrong with you but I suppose sometimes people use it inappropriately.

It's not supposed to be used in cases where the doctor can't figure out what's wrong with you (or what went wrong), but they do this all the time.

Neurological means associated with the nervous system. Maybe you mean psychosomatic?

I don't mean psychosomatic. When they don't know the cause of pain (or other sensory issues such is tinnitus, or vision artifacts, but far more commonly, pain) , they'll often just chalk it off as "neurological" or sometimes "neuropathic". This still attributes it to a physical rather than a psychological cause (which would get more blowback from the patient). While these can be legit diagnoses (especially when supported by evidence such as nerve conduction tests, or other corroborating symptoms), they're often (mis)used when the real source of the issue hasn't been found by the standard (often cheaper) test regime.

The problem with many conditions is that the symptoms are often vague and could imply something serious or could be innocent.
Tests also can provide vague results.
And unnecessary blind testing is not encouraged.

There are many issues that hinder diagnoses. If you should ever have two conditions at once, your doctor may even rule out both of them since you have extra symptoms to each alone. "Oh, you have a cough and a painful lower-right-abdomen: Can't be appendicitis, since that doesn't make you cough...".

The point being: When things get complicated, you need to take control of your medical health. Most doctors don't have the time, the depth of experience, and often, even the intellect to work through a difficult case. On top of that, they often don't have the will to deal with the insurance implications of more expensive testing.

 

[Mirai]

It's not supposed to be used in cases where the doctor can't figure out what's wrong with you (or what went wrong), but they do this all the time.

Some do it, maybe.

I don't mean psychosomatic. When they don't know the cause of pain (or other sensory issues such is tinnitus, or vision artifacts, but far more commonly, pain) , they'll often just chalk it off as "neurological" or sometimes "neuropathic".

That's incorrect usage. You get some people who maybe say that or maybe you misheard.

This still attributes it to a physical rather than a psychological cause (which would get more blowback from the patient). While these can be legit diagnoses (especially when supported by evidence such as nerve conduction tests, or other corroborating symptoms), they're often (mis)used when the real source of the issue hasn't been found by the standard (often cheaper) test regime.

Nerve conduction studies would be done very rarely. Often neuropathic pain is diagnosed clinically. NCS would be done more to determine what type of specific deficit is causing the nerve conduction issue or in specific situations.

There are many issues that hinder diagnoses. If you should ever have two conditions at once, your doctor may even rule out both of them since you have extra symptoms to each alone. "Oh, you have a cough and a painful lower-right-abdomen: Can't be appendicitis, since that doesn't make you cough...".

Some people do this, but I've never met a doctor who did that.

(By the way it is possible for appendicitis to make you cough )

The point being: When things get complicated, you need to take control of your medical health.

You should always do that.

Most doctors don't have the time,

Most of the time, they have the time. Medical care saves lives.

the depth of experience,

As opposed to who?

and often, even the intellect

Someone has a chip on their shoulder.

to work through a difficult case.

What's your experience with difficult cases?

On top of that, they often don't have the will to deal with the insurance implications of more expensive testing.

That's your own take and it's incorrect but never mind.

(Managed care is also a thing. And your medical aid or government health programme like in Europe, expects doctors to follow certain pathways/guidelines/flow charts.)

 

[cguy]

Some do it, maybe.

Given that this was my experience, there is no maybe. This is hardly a stretch.

That's incorrect usage. You get some people who maybe say that or maybe you misheard.

Yes, it's incorrect usage - that's entirely the point.

I didn't "mishear" - perhaps drop the gaslighting?

Nerve conduction studies would be done very rarely. Often neuropathic pain is diagnosed clinically. NCS would be done more to determine what type of specific deficit is causing the nerve conduction issue or in specific situations.

That's why I mentioned other corroboration.

Some people do this, but I've never met a doctor who did that.

Lucky you.

(By the way it is possible for appendicitis to make you cough )

The causality would have to be pretty weak for it to be omitted as a symptom on major medical sites (Mayo, Cleveland).

Most of the time, they have the time. Medical care saves lives.

They absolutely do not have time for complexity. Not sure what you're trying to achieve with the non-sequitur.

As opposed to who?

As opposed to someone who has researched the specific symptom combination or condition extensively.

Someone has a chip on their shoulder.

Indeed, but it's not me. You are being extremely defensive and are denying the reality so many people face with medical care when things get complicated. As for the intelligence of doctors - their mean intelligence is higher than most, so it's just slightly right shifted versus the general distribution of intellect. There are plenty of non-doctors who lie on the far right of that distribution, who are far more capable of diving into a specific issue than a majority of doctors.

What's your experience with difficult cases?

I have solved several.

That's your own take and it's incorrect but never mind.

(Managed care is also a thing. And your medical aid or government health programme like in Europe, expects doctors to follow certain pathways/guidelines/flow charts.)

And if a given doctor has an atypical distribution of flowchart decisions they get pressured to course correct. This also means that they are less likely to treat exceptional cases as exceptional cases.