Big Data has brought with it an ideology that’s reshaping how we approach justice, ethics, and governance. David Brooks first named it “dataism” in 2013, and it holds up data as the best tool for understanding and solving social problems. As Yuval Noah Harari and others have argued, dataism is more than an embrace of technology. It’s a worldview that reduces all human experience to quantifiable metrics. Its proponents celebrate it as the next step in human progress. The approach is dangerously simplistic: it ignores the political, historical, and social forces that shape data and how it gets read.
Dataism holds that everything—people, societies, systems—can be optimized through data analysis. It treats information as the basic unit of reality and says society should make its decisions from data, assuming this will produce more efficient, rational outcomes. As Harari describes it in Homo Deus, dataists expect data and algorithms to replace human intuition and judgment in deciding what’s good for people and for society. That reliance on data looks neutral and objective. It’s full of problems, especially for justice and equality.
The Neoliberal Fantasy Brooks Named
Brooks called it “the rising philosophy of the day” in 2013 and listed the assumptions that come with it: “that everything that can be measured should be measured; that data is a transparent and reliable lens that allows us to filter out emotionalism and ideology.” Later that month Brooks followed the column with another headed “What Data Can’t Do.” The assumptions are the problem here. They treat the messy, subjective side of human decision-making as noise to be filtered out, and they turn social problems into datasets for experts with the right tools to fix. That technocratic approach puts efficiency and optimization ahead of justice and equity.
Dataism doesn’t account for the structural inequalities that decide who gets to create, access, and control data. The dataist dream conveniently overlooks how data-driven technologies often perpetuate systemic discrimination and widen existing power imbalances. Believing that dataism can solve society’s problems without touching the structures of oppression underneath them is a neoliberal habit of mind. It treats inequality as a measurement problem and reports unequal outcomes as objective findings.
Capitalism as Data Processing
In Homo Deus, Harari takes the idea much further than Brooks did. He imagines a future where data becomes the final arbiter of value and meaning, and where data and algorithms dominate every part of human life, leaving little room for human judgment or agency. The book frames these scenarios as possibilities, not prophecies, and closes with questions about dataism’s premises, including whether organisms are really algorithms. Taken as a forecast, dataism comes close to fatalism: it implies that resisting it is futile and that we can only bow to the logic of algorithms and metrics. That depoliticizes technology and ignores the ways it can and should be subject to democratic oversight and resistance.
A handful of powerful corporations that profit from surveillance and commodification own, control, and exploit most of the world’s data. Harari warned about that concentration at Davos in 2018, saying that if too much data ends up in too few hands, humanity will split not into classes but into species. Homo Deus, though, describes capitalism as a data-processing system that beat communism because distributed processing works better. That framing sets aside who owns the data and the economic forces that commodify it. It leaves the analysis incomplete and cuts the narrative off from the material conditions of the digital age.
Data systems often replicate and amplify social biases, as critical scholars like Ruha Benjamin have documented.
Ruha Benjamin and the New Jim Code
Ruha Benjamin’s critique of dataism is the corrective that dataism’s blind optimism needs. In Race After Technology: Abolitionist Tools for the New Jim Code, Benjamin shows how emerging technologies, from everyday apps to complex algorithms, can perpetuate white supremacy while they’re promoted as neutral and progressive. Her concept of the “New Jim Code” describes how discriminatory designs encode racial hierarchies, often amplifying the inequalities they claim to address.
Benjamin makes it clear that data is never neutral. It’s always shaped by the social, political, and historical contexts that produce it. The claim that Big Data analytics and AI decision-making systems are objective often masks and legitimizes entrenched racism and inequality. As Benjamin argues, discriminatory design frequently “ignores and thereby replicates social divisions.” Even when tech developers set out to address bias, their tools often deepen it, because they’re built on historically biased data.
Benjamin also points to the epistemic violence in demanding that marginalized communities produce data to prove their own oppression. The demand puts the burden on the oppressed to translate lived experience into quantifiable metrics, which are often decontextualized and stripped of meaning. The constant call for data from these communities reinforces power imbalances and dehumanizes the people already most exposed to technological harm. It’s part of what Benjamin calls the “New Jim Code”: an updated mode of racial discrimination that works through surveillance, quantification, and prediction.
Data Fetishism and Its Consequences
Benjamin’s critiques, along with Shoshana Zuboff’s analysis of “surveillance capitalism” and the late David Graeber’s work on bureaucratic control, show what dataism’s obsession with metrics does more broadly. The presumed objectivity of data naturalizes and justifies the status quo and makes entrenched power harder to challenge. Turning complex social phenomena into narrow quantitative measures flattens human experience and erases other ways of knowing and being.
As more social policy runs on datafication, marginalized communities are asked more and more often to prove their oppression through data. Systems of surveillance and control grow with the demand.
Steven Pinker’s Misleading Optimism
Advocates of data-driven approaches, like the psychologist and public intellectual Steven Pinker, often argue that data can help us overcome cognitive biases and make more rational decisions. In Enlightenment Now, Pinker marshals an impressive array of metrics to argue that, contrary to popular perception, the world is getting better by almost every measure. His optimism has drawn heavy criticism. Critics argue that his analysis plays down how uneven progress is and how systemic inequities persist.
Erik Larson makes the general point that data and its analysis are just the input to the human mind, where thought and insight happen (Larson 2024). Pinker’s sweeping conclusions risk flattening complex social realities into neat, quantifiable trends, and miss what it would take to address entrenched inequalities.
Data Built With the People It Counts
The answer isn’t to reject data but to use it with attention to context. The data justice scholar Linnet Taylor argues for moving past a narrow focus on individual data rights to look at the collective and structural dimensions of data harms and governance (Taylor 2017). That means centering the perspectives and priorities of marginalized communities when data systems are built and deployed. It means subjecting metrics and algorithms to ongoing audits and impact assessments, and being willing to reject or redesign technologies that fail to serve justice. It also means staying skeptical of data-driven claims, even (or especially) when they come from well-credentialed experts like Pinker.
The struggle for justice under dataism calls for a different data science: one grounded in the lived realities of marginalized communities, accountable to social movements, and oriented toward collective liberation. Without it, the numbers often reinforce the inequalities they record. They can help the communities they describe only when someone first asks who created them, who controls them, and who will be judged by them.
References:
- Benjamin, R. (2019). Race after technology: Abolitionist tools for the New Jim Code. Polity Press.
- Larson, E. J. (2024, April 3). Dataism is Junk Science. Colligo.
- Taylor, L. (2017). What is data justice? The case for connecting digital rights and freedoms globally. Big Data & Society, 4(2), 2053951717736335.
- Zuboff, S. (2019). The age of surveillance capitalism: The fight for a human future at the new frontier of power. Public Affairs.