site banner

Friday Fun Thread for May 15, 2026

Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.

Jump in the discussion.

No email address required.

This is probably my last opportunity to waste time complaining about the upcoming exam, and the study material for it in particular.

The second last mock (150 questions, 3 hour timer) I did? Oh fucking hell man. I don't deserve this. Let me illustrate.

[RCT trial findings for agomelatine vs placebo, including a forest plot]

My answer was marked wrong, despite the fact that I was extremely confident in it. I was nonplussed by the claimed "correct" answer. I took it to ChatGPT with a screenshot.

How am I reasonably expected to tackle this one?

You are not reasonably expected to get this from the marked answer, because the marked answer is almost certainly wrong.

Exhibit 2:

A survival chart/graph comparing different doses of disulfiram vs placebo against cumulative probability of remaining abstinent from alcohol.

(It didn't look good for the disulfiram)

My answer was marked incorrect. Why?

They refer to information provided in the text to support their answer. What fucking text???

Yeah, that line is bad. In the screenshot, there is no supporting text except the graph and the answer options.

Final exhibit:

Some bullshit summary of qualitative research on the challenges of handling dementia patients overnight at a care home.

Where can I get myself some risperidone? Unrelated, what about this one?

(The request for an antipsychotic wasn't unrelated, dear reader)

Risperidone: from a lawful prescriber, ideally not as prophylaxis against SPMM-induced psychosis. On the actual question: your answer is very defensible. The stem literally begins:

“To explore the current practices and challenges in night-time care for people with dementia living in care homes…” Your selected answer: “To clarify the challenges that exist for night-time care of people with dementia living in care homes” is basically a paraphrase of that.

I started losing it.

Right. This mock has left me mildly homicidal. To amuse me, please evaluate what proportion of the questions I've shared for review are sane, dubious, or written by someone who should be an in-patient in a psych ward.

Sane /exam-legitimate: 7 questions Dubious/badly-worded: 5 Containment level broken: 4

So the mock’s sanity index is about: 44% sane, 31% dodgy, 25% cursed.

What do I even say? What do I even do? ChatGPT gently dissuaded me from actually killing anyone, and even filing a GDPR request so I could track down whoever wrote this paper and egg their house. All I can do is come here and vent about it.


Bonus question, copied verbatim:

Mrs. Patterson is a 66-year-old woman who has presented with evidence of dementia. She has a personal history of severe migraine and a strong family history of stroke and early onset dementia. Which of the following is the most likely diagnosis? Mrs. Patterson is a 66-year-old woman who has presented with evidence of dementia. She has a personal history of severe migraine and a strong family history of stroke and early onset dementia. Which of the following is the most likely diagnosis? Mrs. Patterson is a 66-year-old woman who has presented with evidence of dementia. She has a personal history of severe migraine and a strong family history of stroke and early onset dementia. Which of the following is the most likely diagnosis?

Unfortunately, diagnosing the exam writer with dementia was not included in the list of options. Shame.


Anyway. Enough kvetching. Back on the grind. An estimated 75% chance of passing, which I'm going to take with gratitude.

I was gonna say that getting GPT to polish your affirm your intuitions wasn’t good evidence, but I read the sample question and now I smell burnt toast.