The Groundwork of Signatrix

We’re often asked if our company name Signatrix holds any actual meaning or is just supposed to sound flashy and hip. It’s both. Let’s have a quick overview of how our name relates to what we do and how we originally came up with it. Let’s start with the latter.

Immediately after its inception, our company was called ‘Cartwatch’, in reference to the first products we were aiming to build. It soon became clear that while monitoring shopping carts, mainly to prevent losses for retailers, was and still is a big part of what our solutions do, that name was underselling our vision. A name that would do justice to it was needed.

Naturally, the founders being familiar with the software market, turning towards high fantasy lore seemed a little trite to them. Luckily, Christoph has a background in another field that is oddly popular with at least some parts of the software engineering population: Classical German Philosophy. The father of German idealism himself, Immanuel Kant, provided the name we would henceforth be known under. In the ‘Anthropology from a Pragmatic Point of View’, Kant introduces the Facultas signatrix. It is the faculty or, put in a more contemporary manner, ability to draw connections between objects: The faculty of labelling.Let’s be real here: Signatrix is first and foremost a tech company. The biggest parallel between our company and the great sage of Königsberg is that, at least in the days of COVID-19, we tend to not get out too much.

Still, Kant’s philosophy is an interesting subject and looking at just what the faculty of labelling is supposed to be and what role it plays in the Kantian system, a subject on which the literature is scant, is a worthwhile pursuit. Luckily it does yield a result that fits quite well with what Signatrix wants to accomplish. One might say that Computer Vision is giving software the faculty of using signs.

Quick Disclaimer:

1. The author does not claim to be an expert on Kant. If you are a Kant scholar and want to correct this take, Signatrix will be more than happy to give you a platform.
2. We are aware that Kant espoused some problematic views, especially in the ‘Anthropology’. Signatrix does not endorse them. We do however endorse dealing with both theory and technology critically.
   

Let’s look at a quote:

“On the faculty of using signs (facultas signatrix) §38 The faculty of cognizing the present as the means for connecting the representation of the foreseen with that of the past is the faculty of using signs. – The Mental Activity of bringing about this connection is signifying (signatio), which is also called signaling of which the higher degree is called marking.” ^{[Louden, Robert B & Kuehn, Manfred, 2006. Kant: Anthropology from a Pragmatic Point of View, Cambridge University Press. (p.91)]}

The ‘Anthropology’ was written for a broader audience than Kant’s Critiques. In it, he deals with everything, which makes humans distinct from other living beings. In the first part, he lays out how humans obtain knowledge about themselves as well as the world, starting with single sensations. The entire ‘Anthropology’ is written from the standpoint of Empirical Realism: The metaphysical status of the objects that are cognised is not questioned. They are just taken as they are perceived.

This fits with our aims, as well as allowing the author of this article to simply refer you to other sources if you want to educate yourself on just what Kant’s critical project of Transcendental Idealism is all about. Signatrix wants to take software ‘into the physical word’. A realist stance, as philosophically naive as some might consider it to be, is quite helpful here.

Back to Kant: Things can become symbols either by:

1. Convention: Think of words, the meaning of which Kant considers to be chosen arbitrarily.
2. A (natural) link between two objects: e.g., a patient’s heartbeat and his medical status.
3. Being completely out of the ordinary: Think classic miracles like walking on water and their relation to a potential divine being.
   

Naturally, the second kind of signification is the one that seems to be most closely related to Computer Vision applications. Much like humans, at least on the view laid out in a sensualist reading of the Anthropology, Computer Vision applications start with raw data. They can then be trained to identify individual objects and start to make links between them. In the case of our market-ready products, they are then able to recognise attempts at theft better than humans would be.

Kant makes it clear that signifying things is not the same as recognising them through concepts, which is what is needed to truly know them (at least at the level of what humans are capable of). There is a difference between seeing that trees somehow look similar and having an abstract concept of a tree. Symbols can be a tool for the intellect (the faculty responsible for conceptual recognition), but they can lead to great confusion as well if the signified object is confused with its symbol. A striking example he gives for such confusion is the strange world of number mysticism or numerology.

The same could be said to be true for AI solutions. If they are trained on the wrong data or used for purposes they are not suitable for, they may as well do more harm than good, if anything at all. Which brings us back to Signatrix. We aim to aid the retail industry on its path to the future. That means we have to choose the right problems to tackle and design our solutions in a manner that aids the intellect of the people working in the retail industry.

‍