[Steve]

Worth a read

http://www.whatdoesitmean.com/index2517.htm

Algorithms -

http://theunexplained.tv/paranormal-podc...athy-oneil

Here’s a podcast from August 1, 2017 with New York numbercruncher Cathy O’Neil on “Weapons of Maths Destruction” – how algorithms may harm us… »

Podcast notes

6 min   - introduction to Cathy O’Neil.
7 min   - the elephant in the room, people becoming aware of and waking up to algorithms.
8 min   - a growing group of algorithms is not being kept track of and becoming weaponised.
9 min   - do clever people know best?
10min  - we should be able to complain about flawed algorithms.
11min  - a quant, quantitative analyst who builds algorithms, hedge funds.
12min  - in 2007/8 sub prime loans came to prominence.
13 min  - hedge fund is a management fund for the wealthy - invest, then hedge the bet.
14 min  - investment banks lost the most money in 2008 crash, selling short, betting against. 
15 min  - an important tool is to sell short.
16 min  - the culture in the hedge funds, a financial system that needed mending.
17 min  - sub prime loans, piles of debt (securities), credit rating agencies and disillusionment.
18 min  - agencies didn’t do their job, led to the uninformed buying poor investments.
19 min  - the AAA ratings scam, insider knowledge, facts that were never checked or verified.
20 min  - how the crash unfolded, its inevitability, inflated housing market, derivative misuse.
21 min  - corrupt underlying mortgages, they knew it would happen, the crooks got away with it
22 min  - why are movies making these culprits out as some kind of hero?
23 min  - logically extreme perspectives, wishful thinking causes a bubble, misunderstanding.
24 min  - uncontrollable credit card debts, no savings.
25 min  - example of bad individual debt, people suffering, financial institutions ruthlessness.
26 min  - the rich getricher, the poor get poorer, squeezing out more money by use of algorithms. 
27 min  - human decision making and judgement is replaced by algorithms removing control.
28 min  - it’s tracking huge numbers of people but data is measured wrongly, hurts people, lives.
29 min  - personality tests used in job applications, you don’t even know results, discrimination.
30 min  - illegal practice under disability act, can mean you fail all tests for job applications.
31 min  - poor practice is hidden behind flawed algorithms, and is happening silently. 
32 min  - no individual discretion is used by the potential employer, and promises not kept.
33 min  - patterns of data dictating the nature of the algorithm, affects the poor disproportionately. 
34 min  - minimum wage workers with no access to representation, death spiral loop, the rich.
35 min  - people working within organisations wrongly think they are doing no harm, racist traits.
36 min  - implicit bias, we are not modelling the algorithms, it’s a black box, wrong conclusions. 
37 min  - the designers choose accuracy over truth, Facebook actively choosing profit over truth.
38 min  - fake news is generated by algorithms, faceless entities, need for accountable algorithms 
39 min  - the need for proof that an algorithm is legal, fair and our demand for the evidence of this
40 min  - so many hidden victims are arising out of secret algorithm abuse and misuse. 
41 min  - companies should be held responsible, it’s their legal responsibility but they are claiming
41 min  - plausible deniability to avoid any responsibility. 
42 min  - flawed algorithms used in the justice, police, education systems, cyber war implication.
43 min  - use of sentencing scores and demographics mean the poor are sent to prison for longer.
44 min  - it’s being used all over the US, and misused by political parties and politicians. 
45 min  - minimising insurance costs through wellbeing programs that short change people.
46 min  - the wearing of monitoring devices as a long term goal for data gathering.
47 min  - parameter settings are flawed, giving power over innocent people, discrimination. 
48 min  - charging certain groups disproportionately, to those who can least afford.
49 min  - types of insurance, unfair premium increases, big data is incompatible with insurance.
50 min  - cultural understanding of Insurance is meant to pool risk, big data does the opposite.
50 min  - it segregates us into categories by risk, are we helping or pricing people out?
51 min  - if insurers apply flawed algorithms to risks, they will wrongly charge higher premiums.
52 min  - as everything becomes centralised, we’ve lost the concept of a safety net, people suffer.
53 min  - vicious circle for groups of people, are unable to escape the data trap and imprisoned.
54 min  - a regulatory system is required and should be demanded, we need to shape up.
55 min  - I am not a number I am a free man! 
56 min  - the careful use of numbers and data is needed in order to educate us.
57 min  - the political fight we face, people are too accepting of the gradual change.
58 min  - we need to be more engaged regarding our algotithmically controlled futures.
59 min  - ways to audit algorithms, company involvement. 
60 min  - the book ‘weapons of math destruction’.
62 min  - website details.