Interesting Links for 25-06-2015

[andrewducker]

• Some context on Tim Hunt's comments on women scientists
  • (tags: women science )
• The Internet That Was (and Still Could Be)
  • (tags: internet communication )
• Can you solve the Three Switches Problem?
  • (tags: puzzle physics )
• What's Really Warming the World?
  • (tags: globalwarming visualisation )
• The "Confederate Battle Flag" has nothing to do with heritage.

Comments

[channelpenguin]

I had the three switches in an interview. They wouldn't let me take the switch plates off and use a continuity tester! The answer rests on an assumption that I say that one cannot make. (I asked were there any assumptions - but more fool me, I didn't state explicitly which). They didn't agree.

[a_pawson]

I once had the idea of removing the faceplate and pushing a 50 amp current through the circuit attached to switch 1, then simply turning on switch 2. If the bulb is on or off when you go to the attic, it was wired to 2 or 3. If it has exploded, it was wired to switch 1.

That or simply turning on switch 1 and going away for 10 years (or at least double whatever the expected lifespan of a bulb is). Nobody said that test was time-bound.

[drainboy]

Presumably the answer is simply to turn on switch one for 1 minute, turn it off, turn on switch 2, then run upstairs. If it's on, it's 2, if it's hot it's 1, if it's cold it's 3.
What assumption is there in that (except for light making heat?)

[atreic]

It makes the puzzle really really about lightbulbs (and probably doesn't even work with modern led lightbulbs) rather than making it about logic and 'on' and 'off' as binary states. Contrast it to the hat puzzle - if you solved that by knowing that blue wool was itchier than red wool it would be annoying in the same way.

[drainboy]

I can appreciate it might annoy others, but personally having to factor in other physical properties of items in the puzzle doesn't annoy me. Possibly because I worked it out, otherwise I might have been frustrated on hearing the answer. Hard to tell.

The fact there is now a type of lightbulb that it conceivably doesn't work on doesn't seem too problematic (to me). All other kinds of lights would work with it and I'm sure anyone picturing the situation to work through the problem, likely wouldn't be picturing an LED bulb.

Of course if you don't visualise things like that to solve logic puzzles, then I can see that being a problem.

However, I don't think logic puzzles are a good test in an interview. Often it just tells you who has previously worked out/read the solution.

[momentsmusicaux]

In 10 years' time there will be a whole generation who have no idea at all that lightbulbs heat up. That definitely does make the puzzle relying on this fact problematic.

[drainboy]

In 10 years' time I would totally consider this an anachronistic puzzle, unless you started it "afore you were born, you young whipper snapper".

[kalimac]

Instead, the hat puzzle is annoying in a different way. Why does A say the bizarre "I cannot tell what color may hat is" instead of the logical, sensible, and helpful, "You guys are both wearing blue hats"? Or taking off his own hat long enough to look at it?

[theweaselking]

Three logicians walk into a bar.

The bartender asks "Do you all want beer?"
The first logician says "I don't know."
The second logician says "I don't know."
The third logician says "Yes."

[naath]

Well, it makes it not a classical logic puzzle, so it's misleading to frame it as one...

The assumption is the type of lightbulb being used. It wouldn't work with an LED one (maybe if you ran very very fast up the stairs and had very temperature sensitive fingers).

[drainboy]

I can see why I personally might normally categorise this as a logic puzzle, but where does it say in the article or Kirsty's comment that it is a logic puzzle? Just that it is a puzzle (Kirsty might well have been presented with it as a logic puzzle, obviously). I doubt anyone who considered using the heat option, would turn it down and not mention it in a discussion just because there exists a class of light that it wouldn't easily work for.

I also wouldn't place much weight on someone getting or not getting it right, more how they verbally worked through the problem (and possibly how they dealt with being told an answer they disagreed with).

Personally I'd give more kudos to people who gave answers outside of the heat based one, much like I love the (probably apocryphal) story ascribed to Niel's Bohr (http://www.ideaconnection.com/blog/2008/10/a-story-about-a-physics-exam/).

[simont]

I can see why I personally might normally categorise this as a logic puzzle, but where does it say in the article or Kirsty's comment that it is a logic puzzle?

The first few sentences of the above-linked article: "Strangers often ask me to challenge them with mathematical brainteasers. [...] My stock reply is to pose them the Three Switches puzzle."

I grant you that that doesn't explicitly state that it's a mathematical brainteaser of the type requested by the strangers in question, but it's pretty clear that that's what posing it in that context (or telling us that the author does so) is intended to imply!

[cartesiandaemon]

Yeah, that.

I think it's actually a really good puzzle -- I've heard people complain when the puzzle-giver has cheated, and it's true it will go out of date as LED bulbs become more common, but I've never heard anyone object to the standard answer. And I think it's a good example of "it seems impossible because of an assumption you're not conscious of making, but when you think it through, you realise you can drop that assumption". And it's really natural to assume the light is on/off (for everyone, though more so if you're used to electronics or logic puzzles), but also, basically everyone is going to know "lightbulbs get hot" when they consider the solution.

But it's the OPPOSITE of a maths or logic puzzle, but sometimes masquerades as one -- maths or logic puzzles are where you DON'T question the stated assumptions, even if you need to give up assumptions about how to find a solution.

[simont]

Yes. I think atreic hits it on the head elsethread: there are plenty of maths puzzles which are precisely designed to have a solution requiring just the right amount of ingenuity within the explicitly stated parameters of the problem, and lateral-thinking your way out of having to find that answer by inventing loopholes of the form 'ah, but you didn't say they couldn't hide extra prisoners in the courtyard, and if I'm allowed that then I can do it in a really straightforward way' is clearly cheating. And yet the solution you're expected to recognise as right in this case is exactly the kind of thing that would be obviously disqualified from those puzzles.

[cartesiandaemon]

Yeah. I often try to tell puzzles a bit iteratively, describing the puzzle and asking for any solutions, and then saying "and can you do it if you can't do that"? So if someone doesn't immediately guess the level of abstraction expected, they can calibrate, rather than just have me say "no, that's wrong".

[drainboy]

Ah, fair point. I searched the text for the word logic before posting, but not mathematical.

I still think that logic puzzles can include questions that require some reasonable domain knowledge, not just wording that can solve the question after being turned into predicate calculus, but I can totally see why other people might not agree with that.