Marc van Meel

Marc van Meel

Hi I’m Marc!

I speak and write about the intersections of technology, society and philosophy.


Why ‘AI for Good’ is Evil

I once participated in a hackathon to combat poachers with the powers of data science. We had trained a model to detect poachers by movement patterns of herds of animals. Sadly we didn’t win, but at the time that mattered little to me. After all, we were applying data science for the powers of good! These days there exist many projects under the ‘AI for Good’ banner. Initiatives which apply AI in an attempt to solve complex social problems, such as poaching, crime and climate change.[1][2]

There is a problem with AI for Good.

I will critically explore the shortcomings of AI for Good, its conflation with ideology, and why it keeps problems intact. But first it’s time for a coffee!

A man walks into a bar and asks for a coffee without cream.

The waiter replies: “Sorry sir, but we have run out of cream, we only have milk.

Can I bring you a coffee without milk instead?”

~Slavoj Zizek

A Man Walks into a Bar

Explaining a joke kills it, but the meaning behind the joke about coffee is the concept of determinate negation. Meaning that what something is not, is often more important than what something is. Materially speaking, coffee without cream and coffee without milk are the same thing, plain coffee. But they are not the same. Something of importance has been negated. The waiter rejects the ideology of the customer and instead fills in the blanks with his or her own ideology. What you don’t get is part of what you do get.

AI for Good, by analogy, sells you something which you don’t see at first glance. It is implied that you are combatting social problems, but at the same time you are buying into the ideological narratives of technosolutionism and technocapitalism. The abstract label ‘Good’ depoliticizes and masks the root causes of the same problems that are attempted to be solved. Large social problems cannot be solved by technological solutions alone. By doing so, the root causes of these problems stay intact; corruption, poverty, privatisation, etc. The idea that markets and technology can solve every large social problem does not lead to real social change. The way these problems are perceived is part of the problem itself.

Ideology is at work when we know something to be true, but pretend not to know it. We disavow the content of our own actions to remain consumerists in good conscience. And by doing so problems persist. We want sustainability, but we don’t want the bitter taste of actual change. We rather buy fair trade coffee.

Paradise Lost – The Problem of Ideology

AI for Good is supplemental to technosolutionism and technocapitalism. It contains the truth of ideology, but it is not the ideology itself. Instead it is the manifestation of ideology in practice, that which allows ideology to sustain and propagate. Ideology originates from the gap between a problem and an envisioned ideal. It then elevates the proposed solution into a principle. But by overfixation on the solution, upholding it even when it fails, the problem is sustained as well. In every solution we find another problem. The danger of ideology is the suppression of this subjective experience, in favor of utopian thinking.

The 17th-century English poet John Milton does an excellent job at illustrating the problem of ideology in his poem Paradise Lost.[3] Paradise Lost tells the story of Satan’s fall out with God. Satan is described as the embodiment of radical rationalism. His ideology is a rigid belief system which encompasses all of reality. Through him we discover the true nature of evil. Evil as the proclamation to be all-knowing and being unable to admit wrong. Satan will and can never admit error. This blindness leads Satan to be banished to hell. Milton describes hell not only as a place, but also as a state of mind; a place of darkness in which the lurid flickering light of fire serves only to make more dark.

There is no Red Pill

The popular antidote to ideology is that of the metaphorical red pill.[4] The claim is that we should try to dispel the illusion of ideology. That which clouds us from an objective reality, like Neo in The Matrix. But this claim couldn’t be more ideological itself. Just like ideology does, the whole claim of the red pill rests on the presupposition of an ideal world. Ideology is the perceived problem here and breaking into this objective world is the proposed solution. However, this way the problem of ideology is transferred back into the solution itself.

So what can we do? Please don’t mistake negation for negativity. I don’t mean to treat AI initiatives that attempt to solve social problems, and the people involved, with cynicism. Given the impact of AI technology one can even argue that we have a moral responsibility to do so. However, in doing so we should be honest about what AI for Good is, but also what it is not. This means being honest about its conflations with ideology. Especially if these ideologies, and their associated organizations, can be held responsible for the social problems in the first place.

We should challenge ourselves not to succumb to the phantasy of an ideal world. Not by abandoning ideology, there is no red pill to take, but by evolving it from within. This involves confronting the inner unraveling of problems though our subjective experiences. So next time you are presented with a coffee, ask yourself: what was it that I didn’t get?


  • [1] McKinsey Global Institute – Applying Artificial Intelligence for Social Good
  • [2] AI For good UN Summit
  • [3] Milton, John, 1608-1674, Paradise Lost
  • [4] Red pill and blue pill – Wikipedia

More ideas

One Response

Leave a Reply

Your email address will not be published. Required fields are marked *