Hedging for doubt-mitigation

Unsure about something and want to hedge your bets about being proved wrong? Try my simple technique.

‘ Find someone who believes the same thing as you, but more strongly – and then take a directly contrary position. ‘      

How it works :

Let’s say that you believe that there’s a small but real chance that your home will be struck by a rogue meteor and destroyed. You’ll want to insure against it. Let’s also say that you’re lucky enough to come across another individual who is absolutely convinced that there will soon be an imminent meteor strike. If you can arrange a wager against him/her, one in which you bet that (contrary to your hunch) there won’t be a strike, they’ll give you very good odds – because they feel sure you’re wrong. To put it another way, your bet against them will be cheap. So, if there isn’t a strike you won’t have lost much – and if there is you’ll be very well compensated because of the long odds they gave you.

Democratic Refurb

Pick a few democratic countries at random and ask the question “Are they run by jerks?” If the answer is ‘Yes’ (more times than would happen by random chance) then something is clearly wrong with the standard democratic electoral system.

There are any number of very well established ways this might happen. Some examples: Paying for media support (via PR agencies or directly). Blackmailing opponents. Bribery of opponents. Threats of violence towards opponents. Bribery of voters. Scaremongering. Vote rigging. And so on. All tried, tested and reliable ways to maximise votes.

What can be done? I suggest a new method based on a little-used branch of Auction Theory.

Straightforward auctions often fail spectacularly (from the buyer’s point of view) because interested parties are set to bid against each other – so pushing the price far beyond what most would consider ‘fair’. But a variation [I haven’t been able to find its technical name yet] makes a brave attempt to find the ‘real’ price for any item – because the auction is won by the second-highest bidder.

This method stops buyers over-bidding – because they know they’ll lose out if they are the highest. On the other hand they don’t bid too low either, because they’re aiming, as near as they are able, to hit a fair rather than exaggerated price.

Why not apply the same strategy to elections? The ballot would be won by the individual (or political party) with the second-highest number of votes. All the coercive strategies mentioned above would be pointless – they would decrease the chance of a jerk (or bunch of jerks) winning.

Of course the new paradigm would fail if all the potential candidates are corrupt egomaniacal jerks, but I’m betting that, despite appearances, their numbers (as a percentage of all humanity) are actually quite low – it just seems high because the current system favours their high-profile success. You might even say it guarantees it.

Time for Democratic Refurb?


Quantum uncertainty of spinning coins

If one spins a coin on a surface, it can’t be said to be either ‘heads’ or ‘tails’. The coin is in a quantum ‘superposition’ state, where it’s both heads and tails at the same time.

However, when I reach out and grab the coin, it ‘collapses’ its quantum uncertainty, and instantly becomes either heads or tails.

I propose that this previously overlooked phenomenon could be harnessed to build quantum computers based on ‘Coinbits’.

Since the spinning coins can be in both states at the same time, a logic array made of spinning coins would be able to compute all possible outcomes simultaneously – millions of time faster than current binary computers.


Big data? Meh!

You can make money out of Big Data. Or, more accurately, you can make money selling the idea that you can make money out of Big Data. What you can’t do with Big Data is use it to discover Big Effects That No-one Noticed Before.

Some examples.

You can use Big Data to find out that in the UK hummus sells-out in supermarkets much faster on sunny days (people buy tubs to take out on picnics). Big Data’s problem is that the supermarkets know that already. Long before networked point-of-sale computerisation existed, shop owners could just look at the shelves and see what was happening. If you’re a serious hummus manufacturer, retailer or supplier, you’ll already be keeping a keen eye on next week’s weather forecast.

So, rather than use it to confirm things that those in the business (whatever the business is) already know, then alternatively, maybe you can use it to reveal minuscule obscure details that actually, no-one did know before. Like 3% of Toyota drivers who live in detached houses and who also own Nespresso machines prefer elasticated box-bed valences.* But, exploiting that fact to make significant money is likely to be a challenge.

Big Data’s raison d’être is that it’s a fabulous resource if you’re trying to find a needle in a haystack – like, say, tracking-down (or perhaps I should say ‘targeting’) an individual. But of limited use identifying haystacks for you. You knew they were there already.

* A guess

The pleasure of messing things up

The Tay chatbot débâcle did more than demonstrate the naïveté of its developers – it also showed us why driverless cars won’t appear on our roads anytime soon.

Although Tay’s parent company are now employing novelists, playwrights, tv writers and even poets to give ‘personalities’ to their bots, they might do well to put a few psychologists (esp. child psychologists) on the payroll too.

The bot was pulled offline within hours of its launch because interacting users had decided what fun it would be to break. The same thing would happen if driverless cars were to be launched onto our current road system in any significant numbers – they’d be run off the road en-masse by drivers – just for the fun of it (especially if empty).

I’ve heard rumours about experimental research into this very problem, but haven’t been able to track it down.

The solution?  I can only think of one – driverless vehicles would have to have their own exclusive lanes at all times – which would in effect mean duplicating a country’s entire road system.



The Leadership Delusion

“A while back, we did a very simple little study. We asked students at the beginning of the year whether they thought they were a good leader. Then, at the end of the year, we asked who they thought, amongst their number, actually was a good leader. The findings were striking. Those who thought themselves to be leaders were least likely to be chosen as leaders. Why? Because a fixation on the self got in the way of learning about the group and being able to represent it.

–  explain two management professors in The Pyschologist.

An interesting explanation. There is another one though. Maybe the ones who were convinced they’d make good ‘leaders’ were also a p.i. the a , and the others in the groups could spot it a mile off.

In other words, one of the criteria for becoming a ‘leader’ is the talent to be permanently oblivious to the fact that you’re a p.i. the a, (and that that’s how others see you).

‘Leaders’ often end up ‘leading’ simply because others don’t want to deal with them in any way, and in the process let them get away with just about anything.

Have a look at the news about our current crop of ‘leaders’ (especially the corporate ones) and see if they qualify as a p.i. the a or not.

fMRI raises its head again

When a paper appears in ‘Nature’ you have to pay attention. But the current cover-story, which features a new study that maps  hundreds  thousands  of individual words to specific brain areas, down to an accuracy of 2mm x 2mm x 4mm, stretches fMRI credulity to the very limit.

I’m sorry, but I just can’t seriously entertain any study which has entire semantic categories – well, quite frankly, missing.

“One of the missing categories is related to boating and the sea (containing words like “sinking”, “stern”, “boat”, and “diving”).

The other missing category contains character names from 19th and 20th century British novels that are included in the text corpus (“nickelby”, “poirot”, “marple”).

Quantum Computing explained (ish)

Maclean’s (“Canada’s only Current Affairs Magazine”) has challenged seven physics experts to explain quantum computing to the rest of us.

Particularly intriguing is the anti-Panglossian approach – where :

“[…] all the possible ways to get to the wrong answer interfere with themselves and cancel each other out.”

[Note that for any given problem, the number of All Possible Wrong Answers (APWAs) is of course infinite.]

Have a read though the explanations and see if you experience so-called ‘Quantum Dithering’ – where the subject is both illuminated and obscured at the same time.

“It is raining” – but what does that really mean?

Staightforward though it might seem, a simple three word statement such as “It is raining” could have a myriad of meanings and implications – according to a research article from the Centre Nalionale de la Recherche Scientifique in Paris.

The paper, published in Linguistics and Philosophy, examines the phrase in great depth – over 65 pages in fact.

Other phrase under scrutiny were :

‘ Everywhere I go it rains ‘

‘ It is not raining ‘

‘ The rain has stopped ‘

‘ In some place or other it’s not raining ‘  and

‘ There is a lion in the middle of the piazza ‘

For more on ‘ the unavailability of indefinite readings for implicit arguments ‘ , go here.