In the fight against the big tech company’s dominant development logic of AI models, critical theory often falls short of going beyond revealing how everything is imperfect. In fact, the critical lens often undresses the technical artifacts to expose the power dynamics, social and personal values, and, overall, the impartiality of any fact taken for granted; it does it so much so that all artifacts are destabilized, broken into pieces with interconnections; it goes further down until everything is but relational; everything is in its stage of becoming and never a discrete object.

This critical exercise is a useful tool against predominant conceptions of ‘what things are’ and ‘how things should be’. It breaks apart a standard rule or process of AI’s construction, and that is generally helpful in challenging notions that we take for granted.

However, there’s tension between critical exercises and design-development-construction processes: the former undress, break, and destabilize; the latter consolidate, lump pieces together, and stabilize. In other words, the two exercises seem inherently incompatible.

The difficulty I find is that I believe both exercises to be necessary, especially in the world where we aim to construct things yet remain ethical actors.

My arbitrary belief on being an ethical being is to be critical; it is through constant reflection and deconstruction of the taken-for-granted positions that one is forced to re-examine the relations among actants; it is through critical exercises one can attempt to discover the violence they commit as they make decisions to pay attention to one set of relations and disregard the others.

Construction (often a series of decision-making processes) is an inherent form of violence, where one reality (or a set of relations) is taken and the rest destroyed. But could it be done differently? Is it possible to have some form of construction that does not destroy but enables the multiplicity of realities? Before we reconcile these two, we must acknowledge the lack of a perfect construction devoid of criticisms. With every construction comes the possibility for us to critically examine it.

For the lack of a perfect solution, we can only strive for better alternatives.

What are they?

James Birdle, in Ways of Being, offered reflections about what has been understood as ‘intelligence’; the EU citizen panel invited the general public to voice their imaginations of what desirable and fair virtual worlds would look like; the Collective Intelligence Project (CIP) proposed the notion of finding ‘effective, decentralized, and agentic decision-making’ processes that would allow individuals and communities to collectively flourish; wayfinding in online navigation studies investigates how users interact with webpages and take into consideration the inherent features and experiences of humans’ interactions with something non-human (the digital artifacts); there are many instances where people question, inquire, and then attempt to imagine a different way to be with technology.

Could we try to understand humans’ place in and relation with the world, and start from there? Could we find/imagine another form of being, intelligent in a different way, existing in harmonious relations with humans and the ecosystem (of which we are a part)?

Before we construct intelligent machines, let’s first examine all the potential realities; at least, we should take a look at all the options in our hands before we decide on which ones to kill off, right?