Monday 30 January 2017

Cathy O'Neill on Social Justice Algorithms

Dr Cathy O’Neill, aka mathbabe, a former Algebraic Geometer turned Wall Street Analyst turned Data Scientist / Activist, has a best-seller and has just been appointed a Bloomberg columnist. Her target is the algorithms used in complex decision-making

Here’s her conclusion:
The irony is that algorithms, typically introduced to make decisions cleaner and more consistent, end up obfuscating important moral aspects and embedding difficult issues of inequality, privacy and justice. As a result, they don’t bypass much hard work. If we as a society want to make the best use of them, we'll have to grapple with the tough questions before applying algorithms -- particularly in an area as sensitive as the lives of our children.
Well, actually the social justice bit is not the most important issue here. First, something about “algorithms”.

Human judgement was first replaced by algorithms in bank lending and insurance. It turned out (apocryphal source) that human bank managers got it right 82% of the time, and the credit algorithms got it right 85% of the time. For even a small unsecured loans book of a billion pounds, generating £250m of new business a year, that three per cent ibuidls over a short time to a steady £7,500,000 a year of extra profit. More than enough to pay for a couple of dozen credit analysts, their computers and some SAS licenses.

Credit algorithms are brutal. No spare money at the end of the month after paying all your bills and feeding the family? Sorry, no loan. A couple of missed credit card, council tax or gas bills? Sorry, no loan either. A CCJ (County Court Judgement) issued to you at your address? Don’t let the door hit you on your way out. Before you mutter something about banks only lending to people who don’t need it, remember that the bank is lending your savings. You don’t want the bank to turn round and tell you your savings are gone because it was loaned to people who needed the money so bad they couldn’t pay it back - do you? OK. That’s clear then.

Most people can’t make their payments on time, or don’t have spare cash at the end of the month, are low-paid rather than irresponsible. People are low-paid because they don’t have the technical skills, education, or professional persona to earn better salaries. They may also lack the neuroses, character and moral defects, dysfunctions and ability to live without much social life that characterise many of the people who do earn in the top decile of salaries. But let’s not go there, and stick with the lack of education and social skills. Those, in the Grand British Narrative of the Left, are class- and culture-biased behaviours, which fortunately cut across race, creed and colour. In the Grand American Narrative of the Good People, it’s all about race, gender, religion, and economic status - because there is no "class system” in America. Cathy O’Neill is one of the Good People, so she’s concerned that the algorithms may have social injustice embedded in them.

Nobody gets too worked up about bank lending decisions because they are based on past financial behaviour and indicators. Those have an obvious relevance to a lending decision. However, what if the bank refused you because it picked up friends on your Facebook feed who were bad risks? Big Data says that in all sorts of ways we tend to act as our friends do, so it might seem relevant to see if we hang out with financial losers. Everyone lurves Big Data because smart and cool and computers. But how is this not the same as the local gossip saying that we shouldn’t lend to someone because she hangs out with losers? Did the banks hire all those PhD’s just to have them behave like the village busybody? (That’s my objection, not Dr O’Neill’s.)

When the decisions are about sentencing, parole, or taking children into the Social Services system, we would like the algorithms to be a lot better than the local gossip. And Good People want the algorithms to be socially-just as well. Here are the points O’Neill makes about a system called Approach to Understanding Risk Assessment (AURA) introduced in Los Angeles, to help identify children at risk.
The conclusions that algorithms draw depend crucially on the choice of target variable. Deaths are too rare to create discernible patterns, so modelers tend to depend on other indicators such as neighbor complaints or hospital records of multiple broken bones, which are much more common and hence easier to use. Yet these could produce very different scores for the same family, making otherwise safe environments look dangerous.

The quality and availability of data also matter. A community where members are reluctant to report child abuse, imagining it as a stigma or as a personal matter, might look much safer than it is. By contrast, a community that is consistently monitored by the state -- say, one whose inhabitants must provide information to obtain government benefits -- might display a lot more “risk factors.”

AURA, for example, uses contextual information like mental health records and age of parents to predict a child's vulnerability. It’s not hard to imagine that such factors are correlated to race and class, meaning that younger, poorer, and minority parents are more likely to get scored as higher-risk than older, richer parents, even if they’re treating their children similarly.
Her concern is that AURA will have too many false negatives, as the sneaky White People With Jobs stay off the radar. The result will be “unfair” treatment of the people who are correctly modelled. There’s a much bigger elephant in the room. AURA is an appalling model, as O’Neill describes:
In a test run on historical data, AURA correctly identified 171 children at the highest risk while giving the highest score to 3,829 relatively safe families. That’s a false positive rate of 95.6 percent. It doesn’t mean that all those families would have lost custody of their kids, but such scrutiny inevitably carries a human price -- one that would probably be unevenly distributed.
In other words, the next prediction from AURA is overwhelmingly likely to be wrong. Why? Do these people not know what they are doing? Well, I have tried using propensity modelling on a rare event, and got the same result: a horrible level of false positives. After checking my work and berating myself for a lack of creativity, I thought the issues over, and realised that this was caused by the rarity of the event and the nature of the facts I had to use. There is no hope of ever getting a decent predictor for an event as rare as child abuse. First, because it’s rare, and second, because it’s kept private, which is O’Neill’s second point in the quote. By contrast, defaulting in bank loans is a lot more common amongst borrowers than you might believe, and happens within a much smaller chunk of the population than “all parents”.

Propensity modelling started in direct marketing, and even models with much worse false positive rates can help improve profits by cutting down the number of mail shots. What’s good for junk mail is not acceptable for families. Propensity models of rare events are wholly unsuitable for sensitive issues around rare events, not because it "obfuscates important moral aspects and embeds difficult issues of inequality, privacy and justice”, but because the model will inevitably be awful.

That doesn’t mean Dr O’Neill needs to find a new line of work. Big Data research exercises are not expensive and in these kinds of cases a negative result can be valuable. Knowing that there is no group of reliable, accurate markers for child abuse can help dispel prejudice and old wives’ tales, challenge professional folk lore and force policy-makers to think about what they can and cannot achieve. Helping children who are found to be abused is and something a caring society should try to do. Claiming that you can prevent child abuse, when you know it can’t be reliably identified or predicted from publicly-availalble facts, is just irresponsible.

And all the Big Data in the world won’t overcome the cowardice that allowed child prostitution rings run by members of minority groups to operate for years, even though the police and social services knew about it. Which doesn’t mean someone shouldn’t do Big Data research, but it does mean that its issues need to be put in context and proportion.

Monday 23 January 2017

The Telling Strangeness of Brexit

On the 23rd June 2016, the British people asked their Government to get them out of the EU. It was a non-binding referendum. If the Government didn’t like it, they could ignore it. After all, the French did in 2005: their rejection of the European Constiution was made to fade away like the morning mist.

But something strange happened. Everyone in the EU treated the British referendum as binding and final. The Liechtenstein Lush didn't call Cameron and say "You're going to fix this, right?" and go on his way with a smug, knowing grin. Nobody said "We must give the British time to come to understand what they might still do". Junk The Drunk did not behave like a statesman, but like a schoolboy who has finally got rid of the irritating kid in the class. The EU could not wait to be shot of the UK: they wanted Britain to trigger A50 then, in June 2016.

It was the British bien pensants who thought they could get round the referendum. Who thought that Parliament would debate, with the help of right-thinking mavens, the meaning of the vote and whether the British people had voted in thier best interests? Whether it would be the Right Thing to heed the vote and leave the EU, or whether they should set the result aside. It wasn't binding after all. Surely no-one would wrench them from the teat of EU subsidies and Erasmus scholarships?

The EU officials, the 27 heads of state, did no such thing. The British were leaving. End of story. They had it all worked out: they threw our clothes out of the window and changed the locks on the doors. The 28 became the 27 and they took schoolboy glee in excluding Britain from their meetings.

Snowflakes think of Brexit as a divorce, and saw the referendum as Daddy and the kids throwing another ultimatum at Mummy so she would quit drinking for a while. What Mummy EU was supposed to do, after a couple of months, was make a handful of serious concessions to Daddy, so everyone could go back to their dysfunctional family life again. This time Mummy shrugged and told Daddy to take the kids and spilt. That's what's upsetting the snowflakes: Mummy doesn't want them anymore. And maybe never had, for many years.

And it's all Daddy's fault that they found out. Mummy EU was a useful socialist counter-balance to the natural free-market, world-trading, worker-exploiting nature of much of British social and political culture. And now the snowflakes are stuck with life under Daddy: and because they know they don't deserve their grants and subsidies, they are scared Daddy will cut back. He will, but not as much as they think. Hence the wailing of the snowflakes, and their desperate signalling to Mummy EU that they love her still and will she please find a nice comfortable job for them somewhere? But Mummy doesn't love them, and hadn't done for a long time, and it hurts, hurts, hurts to find that out. Nasty Daddy!

Thrown into chaos, the Conservative Party huddled down to find a new leader. Then something strange happened again. Did they choose a clubbable Europhile with close relationships to the bureaucrats in Brussels, who might try to finesse the referendum? No. They chose a woman who had suffered six years at the Home Office being humiliated by the European Courts. They chose a leader who would not be accepted by the 27-Boys Club and who understands in her soul why the UK cannot not go on being over-ruled by a bunch of ideologically-motivated judges in Strasbourg.

Teresa May said "Brexit means Brexit" and to prove it in October the government said they would repeal the European Communities Act 1972. Brexit, it turns out, meant independence and sovereignty. The EU is such a totalitarian entity that everything else follows: if you don't accept its Courts and laws-out-of-thin-air, you can't dodge the 17% tax on shoes designed to protect the Italian shoe industry. Declare legal and political independence and you're back with WTO-based trade deals. Under the Most Favoured Nation rule, any concession the EU gives the UK, has to be given to every other WTO member. So Britain is effectively negotiating trade terms for most of the rest of the world. Those trade terms are the EU's raison d'etre, because it is there to protect the national interests of its members. Perhaps there are going to be more surprises in store: perhaps the EU will use the Brexit negotiations to change rules and tariffs it could not otherwise get past those twenty-seven sets of national interests.

This is a job for professionals who understand what's at stake. The rest of us can only cheer or jeer from the sidelines. In the meantime there will be a lot of posturing. Everyone will want to make it look like They Matter. Whereas they don’t. Not Nicola Sturgeon, not the MP for Lower Cokeatington. Angela Merkel and Victor Orban won’t matter much either. But they will all want to get their shots in. The saddest poseurs of all are the Remainers, deep in denial about the fact that the EU hasn’t wanted them for maybe a decade or more.

Those negotiations will be mostly sound and fury signifying nothing. Face-saving for EU bureaucrats. It’s got to look tough for the UK so that the minnows don’t think they can do it as well. I suspect the British team will play their part in maintaining the charade. I suspect that the professionals know most of the answers already.

When the memoirs are written we will find out that the professionals in the EU wanted the UK out, and the professionals in the UK wanted out of the EU. The catch was that there was no way of doing it that was politically acceptable. So when Nigel Farage turned up - a Euro MP who wanted out of Europe - and when UKIP got 13% of the votes in the 2015 Election, the game became playable. That’s the only explanation that makes sense of all the behaviour that looks so strange under the assumption that the EU and UK wanted to stay together.

Thursday 19 January 2017

Ten Rap Songs

There’s a lot of music I don’t pay much attention to: most nineteenth-century romanticism, Mozart symphonies, dixieland, chord-scale jazz, almost anything on ECM except Keith Jarrett, “avant-garde” jazz and classical music that goes plink-plonk, novelty songs and anything by this year’s, or any year’s, version of FKA Twigs. I’ve never quite managed to dismiss rap, even though I’ve only ever possessed two rap albums (Pain Is Love a while ago and Paid In Full more recently). Rap has moments, I just don’t want very many of them. A little goes a long way. Anyway, here are ten songs that I remember and play on You Tube more than once a year.

 

Monday 16 January 2017

Patton vs Any Given Sunday and the Rest of Hollywood on Winning

If you’ve never seen this speech from Oliver Stone’s Any Given Sunday, then take five minutes to watch it. It’s wonderful writing and a tremendous piece of cinema.



I have just one reservation. At 2:50 Pacino says… “In any fight, it’s the guy who’s willing to die who’s going to win..”

No.
No.
No no no no no no no.
Just.
No.

It’s the guy who’s willing to kill the other guy who’s going to win.

But that doesn’t sound so noble.

General Patton got it right. If you haven’t seen this, watch it as well. The key line is at 1:16.

 

"Now, I want you to remember that no bastard ever won a war by dying for his country. He won it by making the other poor dumb bastard die for his country.”

Hollywood has never since uttered that truth.

Thursday 12 January 2017

Introducing Broscience

If you have never watched any of the Broscience videos on You Tube, you are missing something from your Lifting Life.




50% fact 50% magic 100% results


Okay. I didn't get a lot of sleep Tuesday night. I watched this at about 05:00 Wednesday morning and it was hilarious. I thought I'd share it.

I will write a good long think piece on something soon. It's January. I had a massage Monday that left me feeling a bit dizzy Tuesday. I'm going to be early now. You know. Life.

Monday 9 January 2017

Along London Wall


It's January. Who the heck does anything in January except hibernate?

Monday 2 January 2017

November / December 2016 Review

Happy New Year.

Yep. Going back to doing these. I gave up earlier this year and I’m going to blame the orthodontic work, or rather, my reaction to it. I think I felt so self-conscious with that plastic in my mouth that I stopped wanting to go, or having gone, to stay, out. I even quit having supper with Sis because eating had become such a chore.

So….the ortodontics came out on November 1st. Since then I’ve had two haircuts at George The Barber, attended meetings and been to the gym. A lot. When I wasn’t having horrible colds. I lost a week to a cold, and had the laptop at home, so I could work. I blame half-term.

In the gym, I carried on with the pull-downs and pulley-rows to get some strength into my back, and added leg curls and extensions to see if that will make it any easier to, you know, climb stairs . I started really light and took care when the knees twinged. I suspect I may have to re-build tendons rather than muscle.

I saw After Love, American Honey, Nocturnal Animals, Francophonia, Gimme Danger, Paterson, The Unknown Girl, and Through The Wall at the Curzon Soho. I joined the membership scheme, and now I don’t pay silly prices to see the films, only about as much as I would at the local Cineworld. I saw The Peony Pavillion at Sadlers Wells, and booked tickets into next year as well. I took Mum to see The Red Shoes at Sadlers Wells on New Years Eve afternoon. That involved exciting rides on the 391bus to and from Waterloo and thanking the Lord that Caravan on Exmouth Market was open when almost every other restaurant was closed.

I read I Hate The Internet, by Jarett Kobek, Look Who’s Back by Timor Vermes, The Transformation of Bodies by Yuri Herrera, Chronic City by Jonathan Lethem, Submission by Michel Houellebecq, How to Win Every Argument by Madsen Pirie, and books on Brughel and Jeff Koons.

In November, Sis and I had supper at Rules, then the next evening went to see Peter Pan Done Wrong, and it was exactly as silly and hilarious as I had hoped it would be. Then we got colds and didn’t go out in December.

2017: Embrace The Change.