【能源与环境】 | Energy & Environment

赞助稿酬

By Yibai, Jointing.Media, 2026-05-14

Every day, billions of short videos are uploaded, transcoded, distributed, and played. Behind every frame lies server heat, air conditioning, and electrical current. A significant portion of this energy consumption, however, need not have occurred.

We are living in an unprecedented era of video abundance. Yet, as we become engrossed in a beautifully crafted video, a massive, silent environmental cost is being externalized—its root cause is not a lack of technological capability, but a deliberate deviation from efficiency driven by commercial interests.

Avoidable Waste

From the perspective of pure information transmission efficiency, an uncomfortable truth emerges: a short piece of text would suffice for the core information of the vast majority of video content. Text requires less storage, is easier to access, and promotes deeper thought. Even for “hands-free” scenarios, text-to-speech technology is mature enough, rendering video entirely unnecessary.

The reality, however, is the opposite. The same information is deliberately packaged as video, and the same source file is repeatedly uploaded to different platforms. To build their own closed ecosystems, each platform requires creators to adapt, transcode, and store content separately. A single video on five platforms can generate several to dozens of times the storage redundancy.

This is not a technological bottleneck. Video encoding, container formats, and transmission protocols have long been highly standardized. A standards-compliant MP4 file can theoretically be played on any platform. The additional requirements imposed by platforms—non-standard metadata, proprietary copyright protection, independent ad systems—serve a common purpose: not to enhance efficiency, but to erect barriers.

To understand why all platforms choose this inefficient model, we must introduce a classic analytical framework: the Prisoner’s Dilemma. If Platform A opens up for interoperability while Platform B remains closed, B will attract A’s creators and users. Consequently, all platforms choose closure, even though everyone knows this leads to massive resource waste. This is not a mistake; it is collective irrationality resulting from rational business decisions.

The current wave of video-first content primarily serves the commercial goals of platforms, rather than the genuine needs of users.

The core of a platform’s business model is user engagement time. Video effectively captures attention through dynamic visuals and audio, whereas text allows for efficient information acquisition and quick exit. By diluting and elongating information, video essentially aims to occupy time and serve ads.

Users are guided into an inefficient yet highly engaging mode of information consumption. Platforms reap the commercial benefits, while society bears the environmental costs—this is a classic case of economic externality and a textbook example of market failure.

Quantifiable Pollution

This “invisible” waste corresponds to real consumption of water, electricity, and resulting carbon emissions.

Each video upload is transcoded into dozens of resolutions. Transcoding is computationally intensive; servers consume significantly more power when running at full capacity, with much of the electricity converted into waste heat. Cooling these devices requires an equivalent or even greater amount of electricity. Every playback consumes storage I/O, network transmission, and real-time computation—operations that would be unnecessary under open standards but are repeatedly performed due to commercial closure.

Academia has developed relevant accounting methods. In a 2019 report, the French think tank The Shift Project noted that online video viewing accounted for over 60% of the carbon emissions from global data centers and network transmission. It should be noted that this data is subject to some debate within academia—the International Energy Agency (IEA) has suggested the estimate may be overstated. However, even under more conservative estimates, the carbon footprint of video streaming remains significant. Various studies estimate that transmitting 1 GB of video results in approximately 0.05 to 0.3 kg CO2 equivalent across the full supply chain.

Global video traffic reaches billions of gigabytes daily. Based on these estimates, daily emissions amount to hundreds of thousands of tons of CO2 equivalent. This magnitude is comparable to the monthly emissions of a medium-sized coal-fired power plant. This is not a minor figure; it represents a growing, hidden emissions giant.

All of this could have been avoided.

Potential Paths to Break the Impasse

Awareness of the problem is only the first step. To truly change the status quo, the Prisoner’s Dilemma must be broken. Among the many potential paths, the most noteworthy is structural separation through antitrust regulation.

The EU’s Digital Markets Act already requires interoperability for instant messaging apps—WhatsApp must be able to receive messages from Telegram, for instance. The same logic can be extended to the video sector: legislation could mandate that large video platforms provide standardized content export/import interfaces, effectively separating content storage and distribution from recommendation and monetization. This is not a pipe dream; the EU and the UK are the regions most likely to pioneer such legislation.

Other paths hold value but have significant limitations: mandating interoperability at the technical standards level (e.g., via IPFS) requires public funding, as no commercial entity has the incentive to initiate it; toolkit solutions for creators are vulnerable to being blocked by platforms; and blockchain-based creator economies remain marginal, at least five to ten years away from the mainstream.

Perhaps the most optimistic breakthrough would be a new platform that proactively announces its openness to outcompete rivals—this is the most likely variable in a rational business environment. However, a more realistic assessment is that the Prisoner’s Dilemma will only begin to genuinely loosen when the annual carbon emissions of video platforms are included in mandatory disclosure requirements.

We have no intention of wholly dismissing video. As a medium, it holds irreplaceable value in education, documentary, and the arts. However, we have reason to question whether the push for video, cloaked in “convenience” but actually serving commercial lock-in, is worth such a significant environmental price.

Every meaningless transcode, every instance of redundant storage, every unnecessary transmission represents real consumption of coal and water. These costs have not vanished; they have been shifted to invisible places—beyond the electricity meter, into the atmosphere, and onto the climate bill of the future.

Technology could be more efficient. Business could be more responsible. The Planet should not have to foot the bill for the Prisoner’s Dilemma among platforms.

References:

• The Shift Project. (2019). Climate crisis: The unsustainable use of online video. Paris. (Note: The data in this report is subject to academic debate; the IEA has suggested its estimates may be high.)
• Kamiya, G. (2020). Analysis via CarbonBrief. Cited in: New Scientist. “Binge watching isn’t as bad as we thought.”
• Carbon Trust. (2021). Carbon impact of video streaming.
• Ministry of Ecology and Environment, National Bureau of Statistics of China. (2026). Announcement on the Release of the 2023 Electricity Carbon Emission Factors. National Energy Administration.
• Afzal, S., et al. (2025). SEED: Energy and Emission Estimation Dataset for Adaptive Video Streaming. IEEE VCIP 2025.

---

Image: Artwork by Chen Ya

Edited by Wind

Related:

JM Review | The environmental ledger concealed behind every AI question-and-answer exchange

JM Review |When AI Giants Grow Like Sequoias, Everything Else Withers

Algorithm as Power, it Demands a Cage

JM Review | On ‘315’: The Ethical Dilemmas of China’s Internet Enterprises