What I’m trying to do
I try to understand the underlying concepts or principles behind technologies, ideas, situations, people, everything.
The ultimate goal would be to have a complete understanding, kind of how the enlightenment is described in buddhism. However, I don’t believe that buddhism and meditation are sufficient ways to achieve this kind of understanding. I believe enlightenment as it has been experienced by few humans is a true spiritual experience, brought about by the specific capabilities of our brain to auto-hypnose itself and produce a reality simulacrum, a vicarious universe.
So a complete understanding is both a philosophical and scientific endeavor. Philosophy is crucial, because it lays the basis of the possibility of knowledge. Philosophy allows the critical examination of all other aspects.
Science is the main method of exploration, because it is so far the best way to assert that something is wrong, even though you can’t assert that something is absolutely true.
Finally, there is a self- reflection aspect in this quest: the need to know oneself. Walking in the path of Montaigne, if I may so presume. Why? Because knowledge is situated, localized. Who I am, in all aspects is relevant to what I know, how I know it, and even more all that I don’t or possibly can’t know.
My inquiry and curiosity aren’t directed toward any specific object or domain, rather I’m trying to build bridges between disciplines and spheres of knowledge, because my vision of such a complete understanding can only be holistic, not a collection of separated silos, but deeply integrated objects, ideas and their relationships. A complete understanding is systemic by nature.
The risk is the shallowness / superficiality of knowledge when you try to embrace such a broad range of topics. It is my main fear. I’m utterly aware of how much I don’t really know about so many things. I feel so humble and ambitious at the same time: humility in front of the giants of knowledge and wisdom, humility in front of the myriads of valuable books I will never have time to read, ambition in the goal I set to myself, and that I don’t want to give up, despite acknowledging all my flaws and limitations.
How to resolve this tension?
Can I build anything valuable on so vague notions?
At the same time, I do try to understand underlying principles, not just the stuff of tabloids headlines. This is where critical thinking is essential, to validate as much as possible every idea.
One of my default thinking modes is to find the flaws in any statement, or to identify the blind spots in any analysis. This is critical thinking. But it sometimes leads to a kind of paralysis (at least when trying to comment on the internet).
Asking questions is a large part of that.
Where are the blind spots in machine learning?
Prof of machine learning / AI said recent tremendous progress due to big data. But big data isn’t a bijection with reality. AI operates on existing data, because of its availability and the high costs of generating new data sets entirely from scratch.
Where does the data come from? What are the conditions that decided the collection of this data? What kind of technological or human considerations influenced this data collection?
Technological considerations: measurements, anything that can be encoded in 0 and 1 (photos, sounds)
Not smells, not emotions, not thoughts (not yet but starting to get there)
Human considerations: data on a big scale has been collected by businesses and governments. They have done so with different, specific objectives.
Making profit is the ultimate goal of firms.
Maintaining order and power, the ultimate goal of governments.
Not social good, not individual good.
What other kind of data could be collected? What part of the human experience is not represented, or badly, incompletely represented? What part of the environment is not visible in data set?
Should we try to reverse the process? Start with the problem we want to solve, imagine what kind of data we’re gonna need, imagine how are going to generate that data, or where to find it (but not settle for bad proxy data).
How are we going to make sure that we don’t import the existing flaws of human reality into our data set? Or if we can’t avoid importing them (see language visualized through AI showing how misogyny is deeply embedded in language), how can we use AI to identify such flaws and possibly, once aware of their true extent, correct them?
Why this path?
Because I believe you can’t change the world directly, you can only change yourself. Doing so, you’re also changing the world, to an extent, because you are a part of the world yourself. This is also the crux of humility and ambition.
The idea of writing has been with me for so long, and for so long I didn’t like to write, I resisted it. Only now I’m starting to find words, using a second language. Why?
Accumulated experience, self-reflection, quiet mind (not so anxious, not so much torn in anguish by emotions).
Circumstances, leisure, absence of pressure. Doing work I like, I look forward to and find valuable.