If the so-called social robots are still far from supporting our daily lives, we can already wonder about the project they are carrying out. At a time when digital practices consolidate more than ever the Deleuzian analysis of control societies, the temptation is great for social robotics, still very malleable, to seek its economic model on the side of DataCore cognitive capitalism and little scrupulous.
Table of Contents
Towards a “determinism”
As bankruptcies in the sector reveal, the social robot is struggling to find its raison d’être and its economic model outside the laboratories. Since innovation is largely Darwinist today, the market is facing offers of “solutions” for which there is no problem yet. Now, what is the criterion of “artificial selection” on the digital market today? The data.
Nothing, therefore, like a data collector product to ensure its selection on the market. I call this mechanics “determinism” because, despite the illusory nature of technological determinism, techno capitalist bludgeoning in dataist matter, both in technical and financial acts and in words, obstructs the field of possibilities. The ideology according to which there is no salvation apart from data is thus reinforced.
In the wake of the connected speakers
The connected speakers testify in this respect to rather amazing success. However, they are not very different from all social robots of the “home automation” type which fails on the market. On both sides, it involves both exporting functions with low added value, as they are already available on other portable devices such as telephones, and offering the user a new form of human-machine interaction. (HMI) devaluing the hand in favor of the voice. How then can one explain the success of the one and the failure of the other?
The breakthrough of connected speakers is probably due to a strong brand effect (Google, Amazon, Apple) and a relative technical simplicity making their prices “affordable”. But to their commercial success is added a double strategic success: access to the new seams of data extraction that are household life and the voice in itself. Our voice, in a way, speaks without our knowledge: of the propensity to pay the rent late at risk of having a heart attack, via our mental health. The voice data extraction market is thus attracting an increasing number of companies …
Effects not to be overlooked
The transition from manual HMIs to vocal HMIs is far from insignificant because it marks the emergence of artifacts capable of summoning the general register of sentience (ability to have lived experiences) such as the baby seal Paro or the Pleo dinosaur, or the more anthropocentric one of subjectivity, with robots like Spoon, NAO and Pepper.
The ethical consequences of these phenomena which can be described as psychotechnical cannot be overlooked. Indeed, the ability of social robots to stimulate our empathy and generate emotions is now well known. But while the persuasive power of robots is an emerging subject of study, that many digital players use (and abuse) our cognitive biases to maximize the time we devote to their services, it seems more than appropriate to question the future positioning of sociobots on the market.
When experience becomes a function
Today more than ever, UX design (or user experience Design) puts the function/experience couple in tension. In practice, the second does not exist only to serve the first, as the example of the connected speaker’s shows. Their function, as we have seen, has little added value. Their interest seems rather come from the experience of use, namely from verbal interaction.
So it is not so much what they do, but how they do it, that is important. From there to say that the user experience becomes the function of the object, there is only one step that some social robots take. This is the case, for example, of Paro and Lovot whose function is not to achieve anything but to act as a presence, to entertain, to reassure and/or to generate positive emotions such as attachment. Their function unfolds in and through experience, which amounts to saying that it is experience. The question is: to what interests are the generated experience subject?
Give digital a sufficient scale
In recent years, data capture technologies have greatly benefited from massive investments in artificial intelligence. However, dataist capitalism has every interest in extending its scope of capture to the smallest parcel of existence to maximize the number and diversity of the data collected. Digital technologies, therefore, do not just accompany us along a continuum provided by our technical extensions (laptops, tablets, connected objects of all kinds), they generate a significant part of the experiences that give shape to this continuum.
The determinist strategy is therefore twofold: to digitize experiences that have not been digital so far, like running, and to create new experiences whose source is digital itself, from the countless e-commerce sites to ” online games via social networks. Clearly, it is a question of giving digital sufficient scope for it to become a substrate, a matrix conditioning a maximum of experiments feeding diagenesis.
Unprecedented capture of intimacy
Social robots have substantial assets for achieving this dual strategy, in particular, that of digitizing and therefore “dataizing” our social life offline. And we can legitimately fear that sociobots will follow the path traced by connected speakers: their interactive and emotional potential will make them excellent explorers of our privacy and efficient data extractors.
Many sensors can be integrated into socio-bots, transforming the user into “available funds”, to use the Heideggerian expression. From facial recognition to voice or thermal analysis to emotion detection, these integrated technologies contain an unprecedented power to capture the privacy of users. Human beings are gradually becoming transparent and governable.
An eminently dangerous project
In fact, as Antoinette Rouvroy and Thomas Bern explain in a reference article, big data has the advantage of authorizing the development of “statistical probabilistic knowledge for the purpose of anticipating individual behavior, which is related to profiles defined on the basis of correlations discovered by data mining. Based on this observation, the two researchers forge the concept of algorithmic governmentality which they define as “a certain type of (a) normative or (a) political rationality based on the collection, aggregation and automated analysis of data.” in massive quantities so as to model, anticipate and affect in advance possible behaviors. “
The economic project of such rationality is simple: increase control over the act of purchasing by addressing the pre-reflective part of the psyche. Our “double statistic” is used to “elicit the act of purchase, in the mode of the response-reflex to an alert stimulus short-circuiting in individual reflexivity the formation of singular desire. In other words, it is a question of “producing action without training or the formulation of desires”. By betting on the controllability of our pre-reflexive “selves”, dataist predation obstructs what we call in philosophy the processes of subjectivation, that is to say, the processes leading to constituting ourselves as subjects capable of reflexivity.
Living subjectively is about living together as a subject is always multiple and that it proceeds from a mosaic of elements, influences, very diverse experiences. In other words, politics begins with subjective life. To suppress the possibility of deliberation with oneself by addressing the instinctual part of the self is to attack the very heart of politics which begins with and in itself.
How to accept the Other?
Therefore, if living the subject together is prevented, how could we accept living intersubjectively together? How can I accept the other subject, the other simply, if I don’t first experience this process of acceptance, conciliation, deliberation in me? In addition, obstructing subjectivation is to create docility bordering on a “new voluntary servitude”.
The determinist project, in so far as it carries a depoliticizing social project, is therefore eminently dangerous. As Antoinette Rouvroy states: “Algorithmic governmentality is an immune system developed by capitalism to fight against anything that could put it in crisis. Human beings are only able to rebel if they become full subjects. However, technology comes to seize them on an intersubjective level, at the instinctual stage and gives them what they want right away. “
For democratic ethics of techniques
In this context, the fundamental problem does not a concern so much the respect for private life as the defense and active promotion of an individual and collective subjective life, that is, in short, of political life. It is, therefore, the dominant model of the digital economy that must be revised. The individual and collective depoliticization of which algorithmic governmentality is the bearer must be combated by its inverse symmetry, namely an individual and collective re-politicization.
Make no mistake: this is a democratic challenge, not a technocratic one. While representative democracies face challenges of unprecedented scale such as the ecological crisis and the threat of a disinforming populism that many big names in the digital world encourage or, at best, promote through their inaction, the revalorization of citizenship through new forms of political engagement and action outside the polls seem to be gaining ground. This “renaissance”, partly led by a youth who knows itself more than ever in danger, testifies to a desire for politics.
There is a tremendous opportunity to revitalize democracy, through democratic exercise itself. While data proposes to accentuate political bloodshed for the benefit of economic management alone, we can on the contrary actively associate expertise with citizenship and build a common future! How? ‘Or’ What? Using the rigorous tools of technical democracy to regulate the dominant data economy and determine what are the legitimate causes and modes of harvesting.
Faced with this challenge, still emerging social robotics will have to choose sides: embrace what is already being done or develop according to an alternative model where the notion of “robot companion” is not the hypocrite synonymous with “robot spy”. It is in this openness to the alternative that lies the real innovation.