2023_wordmark_svg_white

Mission Briefings

Data is the [Old] New Oil

Who is in the OPEC of Data? What’s the Horse-and-Carriage of ChatGPT? It’s been 17 years since the famous quote. We revisit.

In 2006, British mathematician Clive Humby quipped that "Data is the new oil".

In 2006, it was still cool to say things like "[X] is the new black”, so Clive’s quote didn’t seem too cringe. Dozens of confident blog posts have derided and praised him since. Media and marketing and VC have each had their respective field days with the concept.

I largely think that Humby is more right than not. And admittedly, it feels he was two decades early with his observation [But this essay isn't about that].

We can learn a lot about our current world - both for sensemaking about today, and for plotting a more effective path forward - if, for the sake of this exercise, we give him the benefit of the doubt and play with this idea.

So, sure, Data is the [Old] New Oil.

Why do we say [Old]?

Because Humby said this quote in a completely different world.

Humby said this pre-iPhone. Pre-Twitter. Pre-Arab Spring. Pre-Obama. Pre-AWS. When only college kids could use Facebook. Pre-Cambridge Analytica. Pre-Trump. And, most certainly, Pre-Generative AI. He also said this the same year that An Inconvenient Truth debuted. And certainly long before confused youth started vandalizing paintings and gluing themselves to gallery walls to protest Oil.

Our relationship to Data has changed. Our relationship to Oil too. So it’s worth our time to revisit the metaphor and see what we can learn.

Data is the new Oil. This is bad.

You live in a world that can be understood as polluted with emissions from Data and resource-cursed by Data, both of which are, like physical pollution and resource extraction, asymmetrically distributed across all traditional lines of structural violence.

For example: if you want to monetize an app on Android, your best bet is advertising revenue, not in-app Purchases. On the whole, Android users tend to pay for things like subscriptions and in-app purchases much less frequently than their iOS counterparts. As a result of this, an incentive structure emerges; Android users’ attention span is a higher-value commodity than iOS users’.

The OS adoption gap largely breaks across race and class lines in the US. Much as marginalized communities disproportionately bear environmental externalities from oil production, lower-income populations endure more intrusive Data tracking by default. The maintenance of human autonomy becomes a premium “privacy-as-a-service” offering, and therefore a tradeoff for convenience and access. This creates perverse incentives exacerbating inequality, with intimate Data obtained from those unable to pay instead commodified to efficiently guide commercial interests.

Everything happening to the environmental ecology as the externalities of mass energy usage from oil are also happening to our cultural ecology.

Our every action tracked and quantified and commoditized and sold in bulk erodes our fundamental notions of autonomy and dignity. The questions of whether entities wielding big Data are eroding autonomy outright, or merely revealing it as more limited than we presumed, remain open. But the velocity and complexity of these systems beg for forethought, not reactionary restrictions stifling progress.

People have a right to demand reasonable agency over their digital value (extracted as Data), and the digital air they breathe (as polluted from commoditized behavioral Data).

When algorithms customize and filter information to suit our preconceptions, they fracture the public sphere into different filter bubbles, severely hampering democratic deliberation and collective action on shared challenges.

This mass "Data extraction" takes a toll on the cultural ecology, fracturing discourse and shared understanding much like fossil fuel extraction destabilizes environmental ecology. We have yet to develop the wisdom, restraint, and oversight to ensure Data creates shared prosperity rather than exacerbating inequality, manipulation, and discord.

In this sense, Data is the new Oil.

Data is the new Oil. This is good.

The other half of the story of Oil is that the effective use of energy in the 20th century - especially oil - was an integral part of what allowed us to go from one billion to six billion people, raise the standard of living for nearly all of them, and begin ascending the Kardashev Scale.

Oil is, in this sense, a godsend for enabling the flourishing of the human species. It, alongside antibiotics and the Haber-Bosch process, are why we all - us all - are here today. The story of nearly every facet of the human condition over the last 100 years is largely the story of us being able to utilize more energy and, in doing so, accomplish substantially more.

So if Data is the new oil, that means that we too should expect to see the effective use of Data dramatically improve the human condition over the next 100 years.

The cascading technological innovations enabled by oil accessibility present an apt analogy for the Data revolution we are currently witnessing.

Just as cheap and easily usable oil created mutually-reinforcing advancements in chemical engineering and manufacturing and mass automobile adoption - changing society profoundly in the process - the proliferation of Data (resource), AI (industrial process), and sensors and effectors (physical devices) is spurring its own mutually-reinforcing runaway cycle of innovation across industries.

As the hydrocarbon engines and chemical processes powered by oil catalyzed efficiency gains in manufacturing and transportation that defined the industrial revolution, the self-improving algorithms and predictive and generative models derived from Data aggregation are enabling the ongoing Digital Transformation defining our lives.

Sectors ranging from healthcare to autonomous mobility to molecular engineering and more are progressing into greater automation, customization and insight by capitalizing on newly accessible mountains of behavioral, semantic and sensory information to train next-generation AI systems.

Through thoughtful implementation, Data-driven technologies are resolving long-standing challenges around resource efficiency, productivity gaps, material science, longevity and disease, and quality of life improvements at individual and societal levels alike.

In this sense too, Data is the new Oil.

[Yet noting: it’s neither fair (nor accurate) to say that Data in the next 100 years will drive population changes analogous to what the effective use of energy, heath, and food did over the last 100. For a variety of interlinked causes, global birth rates are falling so precipitously as to be a crisis of their own. But perhaps, if anything, the metaphor breaks more elegantly: the effective use of Data in the 21st Century will be to enrich the lives of the 8B already here, remediate the world we call home, and potentially to welcome in 8T synthetic minds to it. ]

On Carriages and Cartels

If Data is the new Oil, we should be on the lookout for other places where the metaphor stretches, such the Horse-and-Carriage of Data, and the OPEC of Data.

We should be as curious as we are concerned about which sectors, business models and jobs will struggle to keep up in the Data-fueled 21st century economy, much like Horse-and-Carriage transportation was rendered obsolete in the last century not by violent force but rather by market dynamics.

As new modes powered by Data, automation and AI deliver better value, many existing services and commoditized work may be left behind by capital or relegated to luxury niche status.

Just as gas-powered cars surpassed Horse-and-Carriage for their speed, scale and efficiency gains, AI and Data will transform nearly all facets of the knowledge work economy. This will massively expand information accessibility, yet simultaneously jeopardize existing jobs. Careful policymaking is needed to transition the impacted human workers smoothly to new opportunities. The arrival of affordable, competitive, high-quality general humanoid robotics in 2025 and 2026 will do the same to the service and labor economies what Generative AI does to it in 2024 and 2025.

Likewise, an OPEC of Oil seems inevitable.

As we have experienced with oil-producing states in recent decades, concentrations of precious resource control rarely benefit common people's welfare and shared prosperity.

With Data coming to represent power in the 21st century, we must study the lessons from our challenges with OPEC and petrostates closely. These lessons may help guide both national Data governance policies and international agreements around Data flows to prevent abusive monopolies, promote equitable access, development gains and individual rights alike in the coming Data economy.

So who’s in the OPEC of Data? The companies that will dominate via cartel-like behavior will have the following properties:

1. Massive Data consolidation and commodification infrastructure: Much like oil production requires complex extraction and processing equipment, leveraging Data at scale necessitates huge investments in Data warehousing, modeling, commodification, and brokerage platforms. Companies like Google, Meta, Tencent, and Alibaba that already have this infrastructure and can budget billions more for R&D will have unfair advantages.

2. Integration across the Data production chain: The oil industry’s consolidation and vertical integration enabled a few players to control pricing and availability. Similarly, tech giants that own the platforms producing consumer Data, develop value-added AI services based on that Data, and control the delivery channels to sell Data-driven products, have conflicts of interest and excessive influence.

3. Geopolitical backing due to "systemically important" status: OPEC nations leverage oil's economic centrality to constrain production yet reap huge windfalls. Similarly, nations hosting ecosystems of "too big to fail" Data unicorns may shield them from accountability. Huawei’s treatment shows Data power’s geopolitical significance. International accords on Data firm oversight would help, but may prove challenging.

Learning from OPEC, allowing unchecked consolidation of AI R&D talent and control points across the Data production chain poses long-term hazards. Proactive policy, ethics and incentives give Data firms reasons to align with shared prosperity rather than short-term profits are imperative.

It’s been almost 20 years since Humby’s observation. And it’s never felt more apt than today.

One of the defining traits of the 20th century was the effective automatic use of oil to perform work (via the internal combustion engine).

The defining trait of the 21st century will be analogous; the automatic transformation of raw Data into work (via AI systems). Like the physics equations hold, power in this century will be defined by the asymmetric ability to harness that work quickly.

Achieving and controlling this is the question of our time. This is our oil.

2023_logo_small_svg

Request a demo of Mission Control GenOps.

Find out how AI clarity and trust drive mission success for your team.

Screen Shot 2023-08-27 at 3.55.21 PM

Mission Control is The AI Responsibility Lab Public Benefit Corporation.


© 2023-2042

Company

About

Research

Careers

Events

News

Partners