In our hyperconnected digital age, the concept of “race” is one of humanity’s most persistent and paradoxical constructs. Although we share 99.9% of our DNA1—with more genetic variation within so-called racial groups than between them2—we continue to organize societies around scientifically questionable categories. Today’s technological landscape—shaped by artificial intelligence, the industrial revolution’s legacy, and the collective shifts since COVID-19—offers an unprecedented lens for examining how artificial these divisions truly are.
The Biological Fiction
From a scientific standpoint, human racial categories are largely meaningless. The traits commonly associated with race—skin color, hair texture, facial features—are adaptive responses to different environments and make up a tiny fraction of human genetic diversity. Someone from Ethiopia and someone from Nigeria, often placed under the same racial umbrella, might be more genetically distant from each other than either is from someone of European ancestry3. Yet our social systems continue to treat these arbitrary groupings as fundamental divisions.
This reality becomes clearer as genetic testing and global connectivity reveal shared ancestry and common humanity. The irony is stark: as genetic science advances, the categories we invented appear ever more absurd.
Industrial Standardization of Human Difference
The industrial revolution didn’t invent racial thinking, but it systematized it. As societies organized large populations for manufacturing and urban life, human differences became bureaucratic categories—standardized, measurable, and economically functional. People were sorted into demographic boxes that affected access to jobs, education, housing, and political participation.
That industrial approach laid the groundwork for today’s algorithmic systems that reduce complex identities to data points. The assembly-line mentality evolved into the database logic of the digital era, where racial categories are simply fields in enormous information systems.
AI: Automating Ancient Prejudices
Artificial intelligence is both the culmination and a possible remedy of this trajectory. On one hand, AI trained on historical data can perpetuate centuries of racial bias with unprecedented scale. Hiring algorithms can disadvantage minority applicants4, facial-recognition systems show higher error rates for darker skin tones5, and predictive policing tools can reinforce patterns of racial profiling6.
These technologies automate prejudice, embedding historical inequalities in systems that appear objective. The algorithm becomes the new face of discrimination—seemingly impartial while systematically reproducing biased outcomes that human decision-makers created for generations.
But AI also holds potential to transcend these categories. Machine learning systems could focus on individual characteristics and capabilities instead of group stereotypes. They could recognize relevant patterns and ignore irrelevant demographic markers. The question is not whether AI can move beyond racial thinking—it’s whether we will design it to do so.
COVID-19: A Global Mirror
The COVID-19 pandemic was humanity’s largest real-time experiment in how constructed racial divisions play out globally. The virus recognized no racial boundaries—it spread based on exposure, health status, and biological vulnerability, not skin color or ancestry. Yet the pandemic’s effects followed precisely the racial lines our societies had built.
Different health outcomes, unequal vaccine access, and disparate economic impacts highlighted how deeply embedded these categories are in social infrastructure. Essential workers—disproportionately people of color—faced higher exposure and less access to quality care. Higher mortality in many minority communities resulted from social conditions shaped by historical discrimination7, not biology.
The pandemic also accelerated digital transformation, creating new forms of connection and segregation. Remote work expanded opportunity for some while digital divides excluded others. The crisis underscored our interconnectedness even as racial and national boundaries remained stubbornly persistent.
Collective Movements in the Digital Age
Global protests after George Floyd’s death demonstrated remarkable solidarity that transcended traditional boundaries. Social media amplified images of injustice across continents, creating a shared awareness impossible in earlier eras. Protesters from Minneapolis to Melbourne to Manchester saw common struggles against systems of racial oppression.
Digital amplification revealed both power and limitation. The same platforms that spread messages of justice also create echo chambers that reinforce divisions. Algorithms optimized for engagement often promote polarizing content, making productive conversations about race harder even as global organizing became easier.
The Path Forward
We are at a unique inflection point. Never have we been more connected—our supply chains, communications, and pandemics operate on a global scale—yet we still rely on racial categories developed in the 18th century when intercontinental contact was limited and genetics was unknown.
The technologies reshaping our world give us a choice. We can encode historical prejudices into ever more sophisticated systems, producing a high-tech version of ancient discrimination. Or we can use our growing understanding of human diversity and our capacity for global cooperation to move beyond these artificial divisions.
The satirical aspect of human “race” is not only its scientific invalidity but that we have created machines capable of recognizing human complexity while we still reduce people to crude categories. As we design systems that will shape the next century, perhaps it’s time to make our technology more sophisticated than our prejudices.
The question is not whether racial categories are real—science has already answered that. The question is whether we will choose to make them obsolete.
References
- Human Genome Project Consortium. “Initial sequencing and analysis of the human genome.” Nature 409, no. 6822 (2001): 860-921.
- Lewontin, Richard C. “The apportionment of human diversity.” Evolutionary Biology 6 (1972): 381-398.
- Tishkoff, Sarah A., et al. “The genetic structure and history of Africans and African Americans.” Science 324, no. 5930 (2009): 1035-1044.
- Barocas, Solon, and Andrew D. Selbst. “Big data’s disparate impact.” California Law Review 104 (2016): 671-732.
- Buolamwini, Joy, and Timnit Gebru. “Gender shades: Intersectional accuracy disparities in commercial gender classification.” Proceedings of Machine Learning Research 81 (2018): 77-91.
- Lum, Kristian, and William Isaac. “To predict and serve?” Significance 13, no. 5 (2016): 14-19.
- Yancy, Clyde W. “COVID-19 and African Americans.” JAMA 323, no. 19 (2020): 1891-1892.