Tuesday, March 29, 2022

People Sorting and the Next New Normal

 In my college years I would often find myself on a train from London to Eastbourne. I often had work or some other reason to pass through London, but my mum’s place in Eastbourne remained home.

 

The train always followed the same path with the same stops. First would come East Croydon, a scrappy outer London suburb. The stop after that was Gatwick Airport, and it was followed by Haywards Heath, a posh commuter belt town. Lewes remains a distinctive town of culture, while Polegate and then Eastbourne were seaside retirement places.

 

I would pass the time on the train by playing a harmless game in which I looked at all the passengers and tried to guess which stop they would be getting off at. I could usually score about 80%. The Croydon folk were usually young and casual, in a hurry and not really focused on the journey but the destination. Anybody not lilywhite was a near certainty for East Croydon, as was anybody who looked to be poor. Gatwick was even easier; those passengers had bags or uniforms and would look anxiously at their watches whenever the train slowed down unexpectedly. If travelling in a party they would often be short tempered or arguing.

 

Haywards Heath was for the male corporate warriors, smart in pinstriped suit and tie, and behaving as though they owned the train, which after privatisation they possibly did. The Lewes passengers could be a bit harder to spot. They might dress rather quirkily or appear a little disdainful of the situation they found themselves in. Finally, the Polegate and Eastbourne folk (lumped together in my game) were often old and prone to snoozing on the train. Another clue was that many had come prepared for their visit to hostile London, with ample supply of snacks or tissues or reading material.

 

If I was unsure, I could also find clues in the daily paper they read, a habit that nearly everybody followed back then in the days before mobile phones. Those going to Croydon might have grabbed an Evening Standard before boarding, Gatwick passengers had a book or nothing, The Times was a good indicator for Haywards Heath, The Guardian for Lewes and The Telegraph or Daily Mail for those going as far as Polegate or Eastbourne.

 

The game passed the time and was harmless because it had no real consequences, but of course what I was doing was algorithmic people profiling. In meeting up recently with a boarding school peer I have been reminded that I was into algorithms before anybody even knew what an algorithm was. Seemingly I would sit up in bed devising the school timetable under various invented goals and constraints. Since I recall the primary alternative activities to be fondling other kids, being fondled by other kids, being fondled by schoolteachers, beating up kids or being beaten up, maybe devising algorithms was a smart choice. I do know that many of my missed opportunities for fabulous wealth concerned thinking of valuable algorithms and being too lazy or distracted to follow up the idea. Just one example was devising something that could easily have morphed into an online bridge game that could have cornered the market.

 

Nowadays we all know what algorithms are but most of us still fail to spot when we are being subjected to one, whether it be adverts on websites or journey pricing or even the types of retail outlets that spring up where we live. We also know that there is a fine line between intelligent valid use of an algorithm and unjust discrimination. Most days on the news I see a mugshot of somebody caught on CCTV suspected of committing some offence. Day after day the mugshot shows a coloured man in a hoodie. I am so conditioned to expect that image that I did a double-take this week when the footage of somebody suspected of assaulting a senior turned out to be female, white and dressed in clothes that my wife might wear.

 

Should the police focus resources on the districts where people drawn into crime typically live? Yes, even if a consequence is that some races receive more attention than others. Should they then focus on racial subgroups of residents in those areas? Probably not. Should they single out black guys in hoodies irrespective on any specific evidence? Certainly not.

 

Anyway, this week I devised a new algorithm I could develop while travelling on the subway. Gloriously, New York is rapidly turning back into the bustling city it used to be, and the subways are busy again. Happily, we are in a lull in Covid and some restrictions have reasonably been taken away. But it astonishes me that even after two years of experience most people seem unable and unwilling to make sound personal judgments about Covid risk. A restriction is imposed or removed as an event in time. But the risk changes gradually not suddenly. A removed restriction is not an invitation to recklessness but rather flags a situation to exercise personal judgment. But few seem to want to do that.

 

In recent subway journeys, I notice that mask wearing has reduced to about 50%, even though it is still mandated (though not really enforced). This week I have been trying to devise some predictive indicators towards who is more or less likely to wear a mask.

 

It is not as easy a game as the Eastbourne train game, pr perhaps it is but I have not found the right indicators yet and will need a few more weeks to perfect my algorithm. But I can make a start. There are various possible factors to consider as parameters.

 

One factor must be the actual vulnerability to becoming very sick with Covid. I would expect older and immunocompromised people to be more likely to wear masks. From the other direction, people who have recently recovered or recently had a vaccine might feel safer.

 

That leads to the related factor of self-assessed vulnerability. Young people tend to think they are immortal. Macho types like to boast of how they must die of something as an argument not to protect themselves from Covid. Education and awareness and the sort of conspiracy theories people might swallow will also influence this factor.

 

Then there are the civic factors. Some of us will go to great lengths to avoid any possible confrontation with the NYPD or angry fellow passengers, whereas others seem to actively seek it out. Many might feel relatively safe personally but wear a mask to protect others. Sadly, that category seems to be far from universal in the USA these days.

 

Then there are political factors. In certain quarters, mask wearing (or its absence) is seen as a political statement in this polarised country. Some groups from outside the city might want to protest against a prevailing culture here. Others might just what to stick up a middle finger to the mayor or somebody else they feel lectured by. This group overlaps the so-called libertarians, who in most cases seem to me to be people who just want to do whatever they want without any regard for the impact on others. Sometimes I am surprised that driving on the right side of the road has not yet become optional here.

 

A sad final factor relates the value of a human life, whether that of self, family members or others (subdivided into people like us and other others). A good way to measure societal progress is by estimating how human lives are valued. It drives attitudes to health care, law and order, compliance and much besides. Looking at mortality in the US across a host of categories, I can only conclude that this nation is less developed that it seems to think it is.

 

Many of these factors overlap, but it should be possible to develop an algorithm over the coming weeks to help long subway journeys to pass more pleasantly and quickly. I will keep it to myself and try to avoid judging one group or other, but probably fail in that endeavour. And I do know that I will be wearing a mask myself on those journeys.  

No comments: