That is at times the question and the BBC is introducing us to a hell of a tradeoff. The story (at https://www.bbc.com/news/articles/c0kglle0p3vo) is giving us ‘Meta considers charging for ad-free Facebook and Instagram in the UK’, the setting is not really a surprise. On April 10th 2018 we were clearly given “Senator, we run ads” and we all laughed. Congress is trying to be smart over and over again and Mark Zuckerberg was showing them the ropes. Every single time. There was little or no question on this on how they were making money. Yet now the game changes. You see, in the past Facebook (say META) was the captain of their data vessel. A system where they had the power and the collective security of our data in hands. There was no question on any setting and even I was in the assumption that they had firm hands on a data repository a lot larger than the vault if the Bank of England. That was until Cambridge Analytica and in March 2018 their business practices were shown the limelight and it also meant that Facebook no longer had control of their ship of data, which meant that their ‘treasure’ was fading.
So now we get “Facebook and Instagram owner Meta is considering a paid subscription in the UK which would remove adverts from its platforms. Under the plans, people using the social media sites could be asked to pay for an ad-free experience if they do not want their data to be tracked.” It makes perfect sense that under the guise of no advertising, the mention of paid services make perfect sense. This is given to us via the setting of “It comes as the company agreed to stop targeting ads at a British woman last week following a protracted legal battle.” I don’t get it, the protracted legal battle seems odd as this was the tradeoff for a free service. Is this a woke thing? You get a free service and the advertising is the process for this. As such I do not get the issue of “Guidance issued by the regulator in January states that users must be presented with a genuine free choice.” This makes some kind of sense, so it is either pay for the service or suffer the consequences of advertising. And lets be clear the value of META relies on targeted advertising. What is the use of targeting everyone for a car ad when it includes the 26% of the people who do not have a drivers license. There is the addition that these people need to have an income of over $45,000 to afford the 2025 Lexus RX $90,350 which is about 30%. We can (presumptively) assume that this get us a population of about 20%-25%, so does it make any sense for Lexus to address the 100% whilst only one in four or one in five is optionally in the market? Makes no sense does it? As such META needs to rely on as much targeted advertising as it can. And as you can see, The advertising model, known as “consent or pay”, has become increasingly popular. And at some point they were giving the people “But it reduced its prices and said it would provide a way for users not willing to pay to opt to see adverts which are “less personalised”, in response to regulatory concerns.” That is partially acceptable, but I have a different issue. You see, I foresee issues with “less personalised”, apart from gambling sites, there is a larger concern that even as Facebook (or META) isn’t capturing some data. There is the larger fear that some will offer some services and now care about capturing collected data. For example sites outside the EU (or UK). Sites in China and Russia like their social sites that collect this data and optionally sell it to META. You see, there is as I currently see it no defense on this. Like in the 90’s when American providers made some agreement, but some of them did not qualify the stage of what happened to the data backups and those were not considered, when they were addressed it was years later and the data had left the barn (almost everywhere).
There is a fear (a personal fear) that the so called captains of industry have not considered (I reckon intentionally) the need of replacing and protecting aggregated data and aggregated results. Which allows for a whole battery of additional statistics. Another personal fear is the approach to data and what they laughingly call AI. It is hard to set a stage, but I will try.
To get this I will refer to a program called SPSS (now IBM Statistics) so called {In SPSS, cluster analysis groups similar data points into clusters, while discriminant analysis classifies data points into pre-defined groups based on predictor variables.}
So to get data points into a grouping like income to household types, this is a cluster analyses.
And to get household types onto data points like income to household types, is called a discriminant analyses. Now as I personally see it (I am definitely not a statistician) If one direction is determined, the other one should always fail. It is a one direction solution. So a cluster analyses is proven, a discriminant analyses to income ill always fail and vice versa. Now with NIP (Near Intelligent Parsing, which is what these AI firms do) They will try to set a stage to make this work. And that is how the wheels come of the wagon and we get a whole range of weird results. But now as people set the stage for contributing to third party parsing and resource aggregation, I feel that a dangerous setting could evolve and there is no defense against that. As I see it, the ‘data boys’ need to isolate the chance of us being aggregated through third parties and as I see it META needs to be isolated from that level of data ‘intrusion’. A dangerous level of data to say the least.
There is always a downside to a tradeoff and too many aren’t aware of the downside of that tradeoff. So have a great day and try to have a half cup of good coffee (data boys get that old premise)
Lawrence van Rijn - Law Lord to be 