AI doesn't help us self diagnose -- yet!

Saw this paper on another board. From what I gather, the paper finds that ai chatbots don't help patients make accurate diagnoses or treatment solutions (vs their normal / non-ai information sources).

With model assistance, participants identified relevant conditions in fewer than 34.5% of cases, doing worse than controls; care choices were correct in fewer than 44.2%, with no significant improvement over controls. [Credo tldr]

I don't think this means "ai isn't helpful at all" e.g. doctors obviously use it as an aid. And I imagine using a combination of chatbots and other info gathering methods can increase efficiency (if not accuracy).

But the evidence (from the paper) is -- using ai can't replace a good ol' doctor's visit -- just yet!

Testing anxiety relieving effects of crystals

Came across this neat study from another board:

Placebo effects in alternative medical treatments for anxiety: false hope or healing potential?

The setup:

The randomized controlled study ... included 138 nonclinical adults: 70 crystal believers and 68 nonbelievers. Within each belief group, participants received rose quartz or a visually similar glass stone, then followed a 14-day routine. Anxiety was measured before and afterward ...

The findings:

anxiety scores declined among believers, whether they received rose quartz or the glass stone; nonbelievers showed no reliable improvement. Crystal users did not improve more than placebo users ... Prior beliefs strongly predicted perceived efficacy and symptom change; pre-study expectations predicted 87.7% of post-study belief ratings.

(Above from the credo summary)

So not surprisingly, crystals don't inherently reduce anxiety.

Hinokami Kagura (sun breathing)

In Demon Slayer, Tanjiro's special technique is sun breathing. With the skill, Tanjiro is able to momentarily up with power/speed, sending (symbolic?) flames through his body and blade.

Tanjiro does it again😱 Thunder breathing and Hinokami Kagura | Demon Slayer Kimetsu no Yaiba

Breathing techniques do exist e.g. for meditation or sports. But is there an actual breathing technique that comes close to the Hinokami Kagura?

This paper explores a Tibetan meditation technique that could be a candidate.

Check out the credo summary to see how the researchers carried out the experiment and what they found.

But briefly, they found that seasoned practices were able to raise their body temperature through a combination of a breathing technique and mental visualisation.

It's not entirely clear to me why the practitioners do this, but it sounds like raising body temperature is part of the ritual to (supposedly) heighten mental concentration somehow.

Not quite Hinokami Kagura, but cool nonetheless.

Do polygraph tests work?

Came across this interesting paper on polygraph tests.

Researchers examined historical cases of wrongful conviction -- where accused was found guilty, but later cleared using DNA evidence. The researchers looked at whether a polygraph test was used, and if so, whether the test got it wrong (said person was lying, when in fact the person was telling the truth).

What did they find? That "polygraphs produced unreliable conclusions in about 50 to 65 percent of the cases they examined".

More specifically, they found:

• Polygraph said "lying" (but person actually truthful): 46.8% (n = 59)

• Polygraph said "truthful" (and person indeed truthful): 31.7% (n = 40)

• Inconclusive: 8.7% (n = 11)

• Unknown / Unconfirmed: 12.7% (n = 16)

So for this universe (of wrongfully convicted persons), the polygraph test got it wrong much more than half the time.

House MD: mirror therapy for phantom pain

In one of the series' neatest (and most unhinged) scene, House sedates and ties up his neighbor Murphy -- only to cure him from his phantom pain:

Phantom Limb Pain | House M.D..This "miracle" treatment is one of House's most memorable -- given its deceptively simple setup (a cardboard box and a mirror) and the instant and total relief brought.

I always wondered how scientifically truthful it all was, so I decided to look it up.

It turns out the idea has been around since at least 1996.

Does it work? Over the years, it sounds like there's been some back and forth -- with some studies claiming mirror therapy does work, and others saying it doesn't.

This seems to be the latest review study on the mirror therapy. From the credo summary:

The review counted 21 of 32 studies as effective, meaning pain fell by at least 50% or 2 points on a 0–10 scale. Eleven studies did not meet that threshold. Treatment varied in setting, teaching, movement technique, exercise delivery, session length, frequency, and overall duration; most features appeared in both outcome groups. A 15-minute daily session for four weeks was common among successful studies, but was not established as best.

From what I gather, it sounds like yes mirror therapy does work, and is commonly practiced.

But the immediate and dramatic effects on House are well -- a dramatisation.

Microplastics in drinks -- hot tea the worst!

Some researchers analysed various drinks (in the UK) to see which had the highest microplastic levels:

Synthetic microplastics in hot and cold beverages from the UK market: Comprehensive assessment of human exposure via total beverage intake

The worst culprit? Hot tea.

Why? Based on the factors mentioned in the paper, my guess is it's a combination of:

• Heat. Hot drinks increase microplastic transfer (from container etc) into the drink.

• Teabags. Eventhough teabags are cellulose base, they still have plastics in them. And teabags are soaked a long time in the liquid.

• Material of cup. Even paper cups are lined with plastic.

• Plastic lid. Plastic from lids leak a lot into the drinks.

To say, it's not clear if this level of microplastics are actually bad for us.

But if you're concerned, avoid teabags, and hot drinks in plastic-containing cups.

Can AI make scientific discoveries?

One thing I hear often is that "AI will help research, but it can't make original discoveries" (or some variant).

This article New ‘AI scientists’ are improving – but reveal their fundamental limits seems to say that LLM models can help research, but as they are based on language, they lack precision needed to validate fully hypothesis (and make discoveries on their own).

Surely this is all just a temporary limitation? Or is there some fundamental barrier?

(By the way check out credo summary of the article, which automatically pulls out the research papers cited!)

Dogs help reduce cognitive decline?

Saw this facebook post by fitness influencer Dan Go.

Dan cites this paper, and claims that research shows "pet owners who live alone have significantly slower rates of cognitive decline".

He goes on to say (his main claim), that dogs specifically are the best pets for this purpose. Because dogs help us to take walks, socialise, and follow routines.

This all sounds logical enough. However, the paper cited doesn't actually differentiate between pet types. (Respondents were just asked whether they had a house pet or not.)

That means Dan's claim about dogs particularly, are not backed up by the paper he cited. It could of course come from another study he didn't mention.

Also, the paper's actual findings (that pet owners have slower rates of cognitive decline), only applies when comparing between people who live alone without pets, and people living alone but do have pets. Owning a pet didn’t make you better off vs if you lived with other people.

Colonial German descendants: 2% of population, but own 70% of Namibian land?

Saw an Instagram post (by @AidanSimardone) with this claim, which caught my attention.

I guess one shouldn't be surprised by the abysmal lasting legacy of colonialism. But nonetheless I looked into the claim to learn more.

TLDR: the claim is an exaggeration/simplification, but the actual figures/stats are "still bad".

The first article I stumbled onto was this BBC article: "What's the right price to pay for genocide?".

It summarises Germany's brutal 1904-1908 colonial campaign in (present day) Namibia, noting:

As many as 65,000 of the 80,000 Herero living in German South West Africa at the start of colonial rule are estimated to have perished, as well as perhaps 10,000 of an estimated 20,000 Nama.

and the two countries' efforts to account for it:

Since 2015, when Germany formally acknowledged that the atrocities constituted genocide, it has been negotiating a restorative justice deal with Namibia that will set a global precedent.

On the 70% land figure, the article highlights the historical land distribution/seizure:

... most of their land was split into private farms for German settlers. And today most Herero and Nama live either on small overcrowded areas of communal land that were later allocated to them ... .

and specifically on today's ownership:

German Namibians are believed to be the biggest group among the white farmers who own about 70% of the country's farmland.

So according to the article, it's not exactly 70% of all land, but 70% of farmland. And it's not "white Germans" only, but they are the biggest group.

So the original claim is off, but the sentiment remains.

So what are the actual numbers?

I found this piece by "Namibia Fact Check" (an initiative by a Namibian think tank) – that did the math based on Namibia Statistics Agency (NSA) data:

1. It states "previously advantaged farmers" own ~70% of commercial (freehold) farmland.

2. The piece says "Previously advantaged farmers" isn't strictly speaking "white Germans", and says it's not clear what percentage of the population they make up (but it sounds like that data isn't officially available, and it is understood to mean white Germans and South Africans pretty much. From googling it sounds like white make up ~2-6%).

3. And commercial (freehold) farmland constitutes ~42% of total land surface.

And so, "previously advantaged farmers" (basically white Germans and South Africans, that make up 2-6% of the population) own ~a third of total land (70% x 42%).

So yes, the original claim seems to be an exaggeration – but the sentiment remains. The actual figures aren't exactly rosey.

Online scam operators use human trafficking victims?

I was reading up on human trafficking in the region, and I came across this UN report.

I had obviously known about scam operators (in Malaysia and the region), but didn't realise scammers used trafficked victims to run their operations.

Excerpts from the report are pretty horrifying:

... people are primarily recruited to these operations ... through what appear to be legitimate and even elaborate procedures

... On arrival [to the host country], migrants are ... watched over by security guards who are often heavily armed.

... individuals are subjected to the threat or use of force ... in some cases are unable even to move between different floors of the compound ... Reports have also been received of people being chained to their desk.

... Reports commonly describe people being subjected to ... beatings, humiliation, electrocution and solitary confinement, especially if they resist orders or disobey compound rules or if they do not meet expected scamming targets.

... Reports have also been received of sexual violence, including gang rape as well as trafficking into the sex sector, most usually as punishment, for example for failing to meet their targets.

... The traffickers hold an ever-increasing debt over the migrants – claimed to be for costs such as for travel, quarantine on arrival, training and living costs, and performance related fines ... This is a situation of debt-bondage.

... people have attempted escape, including by jumping off the compound buildings or swimming across rivers (for example, from Myanmar to Thailand or Cambodia to Viet Nam), however these attempts often end unsuccessfully, either with death or severe punishment upon recapture.

The report doesn't single out Malaysia specifically, and the above is a regional assessment.

But thought this is of interest, since Malaysia is known to be a "hub" of sorts for scamming operations (and victims).