This episode ventures into what makes behaviour human and to what extent can we understand and predict it with AI and statistics.
He studied data science and psychology, two disciplines that despite their seeming contradiction have much in common and helped him shape Konsolidate.
As Christophe explains, psychology includes many layers of complexity; it is a vast field to explore with influences from neuroscience to social psychology ‘and everything in between’. This is part of what makes it more challenging to measure. Still, as a science, there must be some measurable data at heart to make it universally accessible and, well, measurable.
On the other hand, data science is much more straightforward and, as Christophe points out, a funny term since all science is based on data. But at its core, data science is about applying scientific methods to gain insights and learn in areas often applied to solve specific problems as part of bigger business processes.
In psychology, that is precisely applied: the evaluation and measurement of data to gain insights and measure them. IQ or personality measures are examples of this and, in fact, are based on factor analysis, a statistical technique developed initially by theoretical psychologists.
Christophe puts his background in philosophy and data science to work:
'As Solid is all about personal data, our behaviour and the digital traces we leave online, we want to put that back under the control of the person who owns that data. So my knowledge as a data scientist on gaining insights from data and how to derive meaning from it makes that work much better.’
Furthermore, Solid works with the idea of semantic data, which solves many data cleaning issues a data scientist faces. It also helps to become readable for humans and computers and connect meaningfully in a knowledge graph.
Another important aspect, and a crucial point of where data science and psychology meet in Christophe’s work, is that this data contains a lot of information about you as a person and your preferences.
This can be used to influence you: to nudge you into consent, buy a product or vote for a specific party or leader, for example. This is a risk as it influences the way we act. Konsolidate aims to mitigate that and build solutions that give people back control over their data.
The term ‘influence’ seems relatively neutral or even positive when our behaviour is actually steered in a direction we might not have intended. We discussed the connotations of influence vs. manipulation and how much power can be exercised over others by accessing their data.
How much trust are you willing to give? How much trust do others give you? What if you decide to exploit that?
Christophe reminds us: ‘We are social creatures and part of that means trying to influence each other every day. From trying to take out the trash to getting a better education and trying to cooperate, so we can live together in a better way.’
The gist is that online, the harmful consequences of sharing personal information are so well disguised that we cannot see the potential harm in sharing pictures and opinions on social media. When we discover our data has been used in a way we didn’t clearly consent to. And our trust in social media decreases.
Privacy is an important part of this, but at the same time, we are social beings and want to share things with others about our lives: show pictures of our cats, or the baby, etc., but we don’t want that information to be used against us. Which essentially is the opposite of the 'social' in social media. If anyone had known, no one would have signed up for this in the first place.
‘Where do mind and body meet?’ A question Christophe found odd because, for him, the mind is an emergent property of a functional brain. He explains substrate indifference of information processing as a concept, meaning that the information we supply is independent of its carrier (audio, video, written, etc.). So, in theory, it doesn’t matter if the information is stored and processed on a silicone substrate’s mind or on a watery fatty carbon substrate which makes up our central nervous system.
We further venture into a philosophical questioning of a connection or separation of body and soul to discover what brings life to the body and what makes human behaviour.
Dipping into the realm of AI: What is an artificial mind anyway, and is it possible for it to ‘learn’ like we do as humans, with senses, speech and physical movement to support us? Christophe reckons that, in principle, it is entirely possible, even though we are not there yet and undermines this point with some good examples.
He adds that AI benefits in detecting patterns in vast amounts of data to support (like cancer research on a scale that humans would never be able to do) has had a very positive effect on research results.
Our conversation wraps up with the possibility of predicting human behaviour with statistics, where Christophe explains that statistics are used all the time to improve sales and predict outcomes. Not on an individual level, but on a broad mass of people, those can be used profitably. However, the exact same processes have been employed to manipulate voter behaviour and steer outcomes, so the risks of exploiting can be rather harmful. This is why it is so important to know about those risks and better protect your privacy, as the individual alone can not find sufficient protection online today.