Hacker News Reader

Comment by Ntrails | Hacker News Reader

Are you sure?

If I buy floor cleaner based on a sales brochure describing a safe and sanitary cleaning product, and it turns out to be Hydrofluoric Acid - am I supposed to double check? How? By sending samples to a private lab?

What about software, I can't see the plugin source code, I can't see any tests done, should I have my own internal team test every possible code path before deploying as part of my app (said plugin might leak GPS data and something something idk I'm straw manning).

Essentially "Buyer Beware" is a very limiting ethos, because buyers cannot be experts in everything.

View on HN

If I'm a large institution buying floor cleaner, I'll probably wash a floor or two before buying. If I notice holes being burned in my floor, maybe I'll go with an alternate vendor.

For this device, it's pretty clear how to beware of it for anyone who took stats 101. Put a grenade or a soda can (depending on a coin flip) into a bag. Point the detector at the bag, record the # of times the detector goes off when grenade vs soda can. Do this double blind.

Use this calculator to determine how many times to do this: http://www.evanmiller.org/ab-testing/sample-size.html

Then type the results into this calculator:

http://www.evanmiller.org/ab-testing/chi-squared.html

Don't buy if it says "No Significant Difference".

Total resources needed: 2 soldiers, 1 grenade, 1 coin, internet.

The most important requirement for the testing methodology is double-blindness. Only someone familiar with proper scientific testing procedures would know that, and even scientists who should know better have gotten this wrong. Most people would probably think that single-blind testing would be enough to confirm whether this works or not (assuming you couldn't judge its plausibility on pure physics principles), but you would be amazed at the ability of humans to pick up informational cues unconsciously. The way dowsing works, in general, is through ideomotor responses[1]. The stick moves due to unconscious muscle movements that are amplified by the instability of the rod itself. The muscles move unconsciously based on what the person believes.

Even if the person who is operating the device doesn't know if the grenade is present, they may be able to pick up unconscious physical tells from the person who does (if he is present during the test). When the person who knows sees the tester approaching, they may unconsciously reveal this knowledge through body language that could be picked up by the tester. A similar phenomena occurred with a horse[2].

[1]https://en.wikipedia.org/wiki/Ideomotor_phenomenon [2]https://en.wikipedia.org/wiki/Clever_Hans

This is all true. But on the other hand, this is all knowledge accessible to any college graduate who majored in a STEM field. Any army or other large organization should have at least one or two such people on hand.

    > 2 soldiers, 1 grenade, 1 coin, internet.

You forgot the soda can. :-)

If people will be eating off that floor, or conducting surgical procedures on that floor, then yes: You're supposed to send samples to a private lab, and you would be negligent if you were buying soapy water without checking.

Do you send samples of your dish-washing liquid to a private lab?

When one makes decisions about what to purchase for a large organization, for a product which needs to perform well or people die, one assumes a degree of due diligence. A novel product requires more stringent due diligence than one which is a market standard. If no plausible mechanism of action is purported for a novel product that attempts to do things with several orders of magnitude more efficacy and less complexity than competitors, then yeah, I'm not buying it without independent testing.

If an underling at my restaurant replaced all dish-washing liquid and bleach solutions with a series of small metal figurines taped to the dishwasher and the prep tables, and told me "It cleans the surfaces galvanically", I'm not going to say "Brilliant!", I'm going to say "Brilliant, now prove it before it's allowed near my kitchen."

BOOM

(Let it go @mapt)

What about the lab itself? What if it's fraudulent? Do I need to send reports from that lab to another private lab?

More seriously though, at some point you need to stop. Capitalism is a fun game for adults, and entrepreneurship is a prestigious sport, but when people start dying because someone sold them something built with the purpose of earning him money by making other people die, one should stop hoping for the market forces to bankrupt him and just jail the murderous SOB directly.

Absolutely. The argument is not that he should be free to defraud people, the argument is that credulous buyers may still be liable for negligence charges (and certainly a civil suit and firing) after we prosecute this muppet for something in the vicinity of second or third degree murder / manslaughter by gross criminal negligence, plus fraud.

Maybe.

But if a couple of hundred janitors and building occupants have died from breathing hydrofluoric acid fumes, and someone still chooses to order the continued use of the floor cleaner in question, that someone should in my opinion be charged with manslaughter if not murder.

You're expected to have some knowledge of what you're buying. And at every level above retail, you're paying enough to get some kind of demonstration or trial. And during this, you're supposed to demonstrate or try it and see if it actually works, and works for your organization.

You can't expect a seller to be at the beck and call of every buyer, especially ones who seem unconcerned about the sale.

Reply on news.ycombinator.com