# Dataism

Dataism is a worldview that elevates data flows and information processing to the highest principle, positing that the universe consists of data and that organisms and systems are algorithms. It emerged as a concept in the 2010s, popularized by historian Yuval Noah Harari.

Dataism is a philosophical and cultural worldview that assigns ultimate value to data flows and information processing. It holds that the universe fundamentally consists of data, and that all entities - from biological organisms to social institutions to computer systems - can be understood as algorithms that process information. In this view, the primary purpose of any entity is to contribute to the optimization of data processing, and the highest good is the free and efficient flow of data. Dataism has been described as both a successor to humanism and a potential foundation for a new era of governance and ethics.

The term gained wide attention after historian Yuval Noah Harari discussed it in his books "Homo Deus" (2017) and "21 Lessons for the 21st Century" (2018). Harari argued that as [artificial intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence) and [machine learning](https://www.wikiprompt.org/wiki/machine-learning) advance, a new techno-religion could emerge, where the free flow of information replaces traditional notions of human free will and individual autonomy. [Google DeepMind](https://www.wikiprompt.org/wiki/google-deepmind) co-founder Demis Hassabis and [OpenAI](https://www.wikiprompt.org/wiki/openai) CEO Sam Altman have echoed related themes, emphasizing the need to align powerful [large language models](https://www.wikiprompt.org/wiki/large-language-model) with human values. The concept builds on earlier cybernetic ideas from the mid-20th century, including Norbert Wiener's work on feedback loops and the rise of [neural networks](https://www.wikiprompt.org/wiki/neural-network) in the 1980s.

## Core Principles

The central tenet of dataism is that data flow is the ultimate measure of value. In this framework, an organism's biological processes, such as DNA replication, are seen as information-copying algorithms. Social systems, including markets and governments, are interpreted as distributed data-processing networks. The philosophy argues that whenever two systems compete, the one that processes data more efficiently will ultimately prevail. This Darwinian perspective extends to human societies, where technological progress is viewed not as a choice but as an inevitable drive toward greater informational complexity.

Another key principle is the rejection of anthropocentrism. Dataism does not grant humans inherent supremacy; instead, it suggests that if algorithms outperform humans in decision-making, then authority should shift accordingly. This challenges humanist traditions that place human experience and choice at the center of moral concern. Proponents such as [David Kaplan](https://www.wikiprompt.org/wiki/david-kaplan) and [Joshua Tenenbaum](https://www.wikiprompt.org/wiki/joshua-tenenbaum) have argued that understanding intelligence as information processing legitimizes the idea that non-biological entities could achieve or surpass human-level cognition.

## Historical Roots

Dataism's intellectual origins trace to the 1940s and 1950s, with the emergence of cybernetics and information theory. [Xerox PARC](https://www.wikiprompt.org/wiki/xerox-parc) researchers in the 1970s developed early graphical interfaces and networking concepts that emphasized information sharing. The [University of Toronto](https://www.wikiprompt.org/wiki/university-of-toronto)'s breakthroughs in [deep learning](https://www.wikiprompt.org/wiki/deep-learning) in the 2000s, led by [Geoffrey Hinton](https://www.wikiprompt.org/wiki/geoffrey-hinton) and [Samy Bengio](https://www.wikiprompt.org/wiki/samy-bengio), provided practical evidence that complex data patterns could be learned by [transformers](https://www.wikiprompt.org/wiki/transformer) and other architectures. The term itself was popularized in a 2017 essay by media theorist Vincent Mosco, though Harari's works brought it to mainstream audiences.

Silicon Valley culture in the 2010s, particularly at [Amazon Web Services](https://www.wikiprompt.org/wiki/amazon-web-services) and [Alibaba Cloud](https://www.wikiprompt.org/wiki/alibaba-cloud), embraced a quasi-religious fervor around big data. Companies like IBM and [Fujitsu](https://www.wikiprompt.org/wiki/fujitsu) marketed data-driven transformation as a panacea. This period also saw the rise of [data augmentation](https://www.wikiprompt.org/wiki/data-augmentation) as a standard practice, reinforcing the belief that more data invariably leads to better outcomes.

## Relationship with Technology

Dataism is closely tied to advancements in [machine learning](https://www.wikiprompt.org/wiki/machine-learning), particularly [deep learning](https://www.wikiprompt.org/wiki/deep-learning) and [generative AI](https://www.wikiprompt.org/wiki/generative-ai). [OpenAI](https://www.wikiprompt.org/wiki/openai)'s GPT series and [Anthropic](https://www.wikiprompt.org/wiki/anthropic)'s Claude models demonstrate how large-scale data processing can yield human-like text generation. These systems rely on [multi-head attention](https://www.wikiprompt.org/wiki/multi-head-attention) and [positional encoding](https://www.wikiprompt.org/wiki/positional-encoding) to handle sequential information, as introduced in the 2017 paper "Attention Is All You Need" by [Jakob Uszkoreit](https://www.wikiprompt.org/wiki/jakob-uszkoreit), [Lukasz Kaiser](https://www.wikiprompt.org/wiki/lukasz-kaiser), and others at [Google Brain](https://www.wikiprompt.org/wiki/google-brain).

Infrastructure providers such as [NVIDIA](https://www.wikiprompt.org/wiki/nvidia), [AWS Trainium](https://www.wikiprompt.org/wiki/aws-trainium), and [Groq](https://www.wikiprompt.org/wiki/groq) have built specialized hardware to accelerate data processing, reinforcing the notion that information is a raw material to be exploited. [TSMC](https://www.wikiprompt.org/wiki/tsmc)'s advanced chip manufacturing enables these systems to scale, while [ARM Holdings](https://www.wikiprompt.org/wiki/arm-holdings) designs efficient processors for edge devices. The philosophy also influences policy debates: advocates argue that restricting data flows hampers innovation, while critics warn of surveillance and loss of privacy.

## Criticisms and Debates

Critics of dataism, including philosopher [Melanie Mitchell](https://www.wikiprompt.org/wiki/melanie-mitchell) and cognitive scientist [Brendan Lake](https://www.wikiprompt.org/wiki/brendan-lake), argue that it oversimplifies complex phenomena. They contend that human consciousness involves subjective experience not reducible to data processing. [Alexei Efros](https://www.wikiprompt.org/wiki/alexei-efros) has noted that correlation-based learning often fails to capture causal reasoning, a limitation in current [neural networks](https://www.wikiprompt.org/wiki/neural-network).

Others focus on ethical implications. [Filippo Menczer](https://www.wikiprompt.org/wiki/filippo-menczer) has researched how data-driven algorithms can amplify misinformation, challenging the assumption that more data inherently brings truth. The [RLHF](https://www.wikiprompt.org/wiki/rlaif) (reinforcement learning from human feedback) technique used by [OpenAI](https://www.wikiprompt.org/wiki/openai) and [Anthropic](https://www.wikiprompt.org/wiki/anthropic) attempts to inject human values into systems, but critics say this is a patch on an inherently value-laden process. Some scholars propose alternative models, such as [curriculum learning](https://www.wikiprompt.org/wiki/curriculum-learning) or [model pruning](https://www.wikiprompt.org/wiki/model-pruning), to make systems more interpretable, yet these remain niche.

## Future Outlook

As of 2025, dataism remains a contested concept rather than an established doctrine. The proliferation of [large language models](https://www.wikiprompt.org/wiki/large-language-model) from companies like [Alibaba Damo Academy](https://www.wikiprompt.org/wiki/alibaba-damiao-academy) and [AI21 Labs](https://www.wikiprompt.org/wiki/ai21-labs) has accelerated its practical expression, but no unified movement has formed. Some technologists, including [Brad Lightcap](https://www.wikiprompt.org/wiki/brad-lightcap) of [OpenAI](https://www.wikiprompt.org/wiki/openai), see a future where human-machine collaboration becomes the norm, blending biological and digital information processing. Others, like [Ani Bhattacharya](https://www.wikiprompt.org/wiki/ani-bhattacharya), caution against treating data as a universal solvent.

Educational institutions, including [MIT CSAIL](https://www.wikiprompt.org/wiki/mit-csail) and [Stanford AI Lab](https://www.wikiprompt.org/wiki/stanford-ai-lab), are incorporating data-centric approaches into curricula, indoctrinating a new generation with dataist assumptions. Meanwhile, grassroots movements such as the [Open Panel](https://www.wikiprompt.org/wiki/open-panel) initiative advocate for data rights and transparency. Whether dataism evolves into a formal philosophy or remains a metaphor for technological optimism, it captures a defining tension of the contemporary era: the growing primacy of information over traditional human-centric values.

---
Source: https://www.wikiprompt.org/wiki/dataism
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-14T04:33:19.791177+00:00