Monday, 31 July 2017

Compo Results: & a Brief Summer Book Review

(1) Brief Book Review

The Turing Guide  - 33 contributing authors (Oxford, 546 pp)

Alan Turing was one of the foremost geniuses of the 20th C.  Many will know something, if only vaguely, of his mighty contributions to maths, computing and AI, and to codebreaking during WW2.  This book details all of that and a great deal more, in as much technical / mathematical detail as you can stand: but because it's in 42 chunks (meaty chunks, it must be said), it is also quite straightforward to bite off something digestible on anything that piques your interest.

And there's plenty to choose from (did you know Turing was a pioneer in mathematical biology?)  I could imagine most C@W readers will be more concerned to learn about his WW2 exploits than, say, his 'gay martyrdom'.  Several of the contributors reckon Turing was personally instrumental in shortening, if not actually winning, the war.  Single-handedly?  No: notwithstanding a strong strand of idolisation of the man, the book does not fail to record in detail the efforts and mind-blowing achievements of half a dozen other outright geniuses at Bletchley Park and beyond, most of whom sadly are little-known today.  Riveting stuff.

(2) Compo

A most enjoyable set of car names.  But our Anons must realise - prizes are only awarded to folks willing to give themselves a name too!  So - nothing for the Ferrari Tosserati, nor for the Kia Thornberry.  Instead, it's

  • Bronze (wins a test drive in the Thornberry):  Mr BQ, Renault Smug 
  • Silver (a test drive in the Tosserati) - with special commendation for its accompanying spec - Kev, and his Mercedes Benz Diana
  • Gold:  Steven L, LOTUS FARAGE  - a real guzzler of a motor, roaring along and belching smoke.  (a day trip to Aberdeen)


Bill Quango MP said...

The Lotus Farage is a beast of a vehicle.
Very fast. Quite nimble. But the old fashioned drum brakes means it resists all attempts to make it stop. Not very good at corners either. Prefers the single issue track.

And the noise when it starts up causes liberal bicycle types to run, hands clasped over their ears, screaming in terror

Electro-Kevin said...

I suppose my Mondeo Vaginal Hero (a play on the original tital name) was... too late an entry ?

(Fnaar, fnaar !)

Electro-Kevin said...

Unfortunately the Lotus Farage spends most of it's time in the garage, or is that..gar-aage ?

Steven_L said...

And I reckon the Lotus Farage (officially pronounced fa-ridge for the avoidance of any doubt Kev, not that the punters would take any notice) would have a waiting list as long as a queue of 'Syrian' migrants hiking through Hungary.

Electro-Kevin said...

Especially if it became available on PCP terms.

Electro-Kevin said...

Did I really spell title tital ??? Where TF did that come from ?

Anonymous said...

ND, your attempt to elevate the tone has failed!
Turing was indeed a bloody genius. Ultra/Colossus is an amazing story.

John Miller said...

42, eh?

Coincidence or a sense of humour?

I know what I'd like to believe...

Anonymous said...

He still should not have been "pardoned".

Anonymous said...

@9.32 as opposed to Knighted?

The biological maths never really got off the ground but I heard some Professor type saying that they're only just beginning to contemplate where he was heading.


hovis said...

Turing was indeed a genius. I think though it is necessary to acknowledge his work at Bletchingly without idolising him or getting hung up about his sexuality one way or the other.

His pronouncements on AI and the 'Turing test' in particular are interesting but these days often fetishised and unquestioned by many in the field. This is especially the case with the rise in fashionable thoughts of 'Smart technology' and the push in some quarters of Silicon Valley and beyond to embrace transhumanism.

For views on AI, technology and society I am big fan of Jaron Lanier.

"“But the Turing test cuts both ways. You can't tell if a machine has gotten smarter or if you've just lowered your own standards of intelligence to such a degree that the machine seems smart. If you can have a conversation with a simulated person presented by an AI program, can you tell how far you've let your sense of personhood degrade in order to make the illusion work for you?

People degrade themselves in order to make machines seem smart all the time. Before the crash, bankers believed in supposedly intelligent algorithms that could calculate credit risks before making bad loans. We ask teachers to teach to standardized tests so a student will look good to an algorithm. We have repeatedly demonstrated our species' bottomless ability to lower our standards to make information technology look good. Every instance of intelligence in a machine is ambiguous.

The same ambiguity that motivated dubious academic AI projects in the past has been repackaged as mass culture today. Did that search engine really know what you want, or are you playing along, lowering your standards to make it seem clever? While it's to be expected that the human perspective will be changed by encounters with profound new technologies, the exercise of treating machine intelligence as real requires people to reduce their mooring to reality.”
Jaron Lanier: 'You are not a Gadget'

E-K said...

How sad that our technological zenith has teached a cultural nadir.

(Just trying to help elevate the tone - for once.)

E-K said...


Nick Drew said...

hovis - one of the (many) things I learned from the book is that the Turing Test (or tests - he kept evolving it) was always rather more complicated and subtle than how I'd previously understood it

(not saying yer man Lanier doesn't know this full well ...)

I'd have to take issue with bankers believed in supposedly intelligent algorithms that could calculate credit risks before making bad loans

what actually happened was (a) the quants came up with the best risk algorithm for credit risk that they could - and it was excellent but never 'perfect', and the quants and risk managers always knew, and acknowledged its limitations

(b) regulators accepted it for use, as being 'best we've got' / 'prudent-if-properly-understood', and used in conjunction with other risk-management measures

(c) critically, it depended on 2 things that proved unavailable: (i) really good stuff from the ratings agencies (who were genuinely delinquent, they always are - not to mention highly conflicted); and (ii) credit risk being non-systemic, i.e. credit 'events' being independent, not everyone going bust at the same time for the same reason

(d) a bunch of greedy people seizing on the 'approved' nature of the algo and saying, well, that's OK then, ignore the caveats etc etc and Go For It !

... with the predictable results

i.e. the bankers never 'believed' in anything - except their own bonuses

hovis said...

Nick: I don't totally disagree with what you are saying, but as ever the devil is in the detail.

Whilst quants may have acknowledged algo limitations(*), as soon discussion moves outside quantdom, it is sold as "we can measure risk". Case in point, a company I used to work for was selling risk measures based on Black-Scholes but once in the Sales cycle, limitations were neither cared about or understood on either side. So whilst at the micro level you are correct, in actuality it is as good as saying they did believe.

Currently we are seeing the same hype, lack of precision, wishful and flabby thinking around AI and Big Data and blockchain currently.

I'd still recomend on reading at least a summar of his thoughts on micropayments, data ownership and where technology it taking us.

(*)Quants never were much good at the soft stuff like communication, (hence they are quants in my experience.)

Nick Drew said...

Good points, well made - I shall give this chap a try

dearieme said...

Too late, I suppose, for the Vauxhall Vulva?

Nick Drew said...


(on several grounds)