You Are Already Contributing — How Everyday Actions Become Social Value, and What DEP Protocol Is Building

Prologue: Who Transcribed 17,600 Books?

In 2008, a research team at Carnegie Mellon University published a curious report in Science. Those distorted strings of characters we typed to prove we were "not a robot" when logging into websites—reCAPTCHA—were actually fragments of old books and newspapers that OCR (optical character recognition) software had failed to read. Every time people around the world proved they were human, humanity's intellectual heritage was being digitized, one word at a time.

In just its first year of operation, over 1.2 billion reCAPTCHAs were solved, deciphering more than 440 million words that OCR could not read. That is the equivalent of 17,600 books. Doing the same work through conventional labor would have required more than 1,500 people working 40-hour weeks (Source: von Ahn et al., "reCAPTCHA: Human-Based Character Recognition via Web Security Measures," Science, 2008. https://www.cs.cmu.edu/~biglou/reCAPTCHA_Science.pdf ).

An independent study presented at the NDSS Symposium in 2025 estimated the cumulative totals: at least 512 billion reCAPTCHA v1 and v2 sessions, roughly 819 million hours of human work—about $6.1 billion at the U.S. federal minimum wage (Source: NDSS Symposium USEC 2025. https://www.ndss-symposium.org/wp-content/uploads/usec25-21.pdf — note these are external researchers' estimates, not audited figures from Google).

My point here is not to judge Google. What I want to highlight is a single fact:

For this enormous contribution, the contributors received nothing. And no one could see what they had contributed, or how much.

You and I have already contributed hundreds of times. Our contributions simply were never recorded, never evaluated, never rewarded.

This essay is an attempt to establish—with sources—that the phenomenon of everyday human actions generating social value without our awareness is already happening everywhere; to explain why a future in which those contributions are properly recorded, evaluated, and rewarded has finally become possible; and to describe why we are building DEP Protocol as the infrastructure for that future.

Chapter 1: Invisible Contributions Are Already Moving the World

reCAPTCHA is only the tip of the iceberg.

Waze. The Israeli-born navigation app aggregates users' driving speed, location, and stopping patterns into real-time information about congestion and accidents. The city of Joinville, Brazil used Waze driving data for road improvement planning; according to Waze's official case study, the productivity impact on the city's economy was $1 billion, and travel time on a major corridor was cut by 18 minutes (Source: Waze for Cities Case Studies. https://www.waze.com/wazeforcities/casestudies/using-data-for-roadway-upgrades/ — figures published by Waze itself, not independently audited). Your commute was becoming urban planning data.

Foldit. In this citizen-science project that turned protein folding into a puzzle game, players contributed to solving the structure of an AIDS-related enzyme. Play became scientific discovery.

SETI@home and BOINC. These systems channeled the surplus computing power of home PCs into analyzing signals from space and medical research, with millions of participants at their peak. Idle computers became part of big science.

Vitality. The South Africa-born wellness program rewards the sharing of step counts and exercise data with insurance discounts and perks. This matters as a commercial example where rewards exist: a daily walk is recognized as actuarial value, and compensation is paid.

Boston Street Bump. The city of Boston detected road surface anomalies from the vibrations picked up by citizens' smartphone accelerometers during normal driving, leading to the repair of 1,250 deteriorated manhole covers, with a false-positive rate under 10% (Source: City of Boston, New Urban Mechanics. https://www.boston.gov/departments/new-urban-mechanics/street-bump ). Everyday driving became infrastructure inspection.

And in recent years, networks that build rewards into this structure have emerged: Hivemapper, which updates maps from dashcam footage; WeatherXM, which rewards local weather observation; OpenLitterMap, which turns photos of litter into open environmental data. Citizen participation × rewards × infrastructure is already a global trend.

PicTrée, the location-based game we operate in Japan—where players photograph utility poles, manholes, and other infrastructure—is Japan's implementation of this trend. Our publicly announced work with TEPCO Power Grid and the NTT East group means citizens' photographs genuinely carry part of the equipment-inspection workload. Walking through town for fun contributes to maintaining infrastructure, and that contribution is compensated. The decisive difference from reCAPTCHA is this: the contribution is visible to the contributor, and it is rewarded.

Chapter 2: The People Who Named the Invisible Labor

The first people to name this problem were not from the crypto industry. They were philosophers and economists.

In 2000, media theorist Tiziana Terranova argued in her paper "Free Labor" that the very acts of enjoying, socializing, and creating culture on the internet constituted an unpaid transfer of value to the digital economy (Source: Terranova, "Free Labor: Producing Culture for the Digital Economy," Social Text 18(2), 2000. https://web.mit.edu/schock/www/docs/18.2terranova.pdf ). This was before social media or smartphones existed in their current form.

In 2013, Jaron Lanier—often called a father of VR—criticized in "Who Owns the Future?" how massive digital networks aggregate people's posts, searches, ratings, and location data while concentrating the value in a handful of what he called "Siren Servers." The core of his proposal was not a one-time sale of data. It was "data dignity": every time data is used and generates new value, a small but continuous compensation should flow back to the person it came from (Source: Simon & Schuster. https://www.simonandschuster.com/books/Who-Owns-the-Future/Jaron-Lanier/9781451654974 ).

In 2018, Lanier and economist Glen Weyl, among others, translated this vision into the language of economics. "Should We Treat Data as Labor? Moving beyond 'Free'," published in the American Economic Association's AEA Papers and Proceedings, pointed out that current economic statistics and corporate accounting treat user-generated data as corporate capital, ignoring users' role in production. Data should be treated, at least in part, as labor, with institutions to distribute its compensation (Source: Arrieta-Ibarra, Goff, Jiménez-Hernández, Lanier, Weyl, AEA Papers and Proceedings 108, 2018. https://www.aeaweb.org/articles?id=10.1257%2Fpandp.20181003 ).

By 2023, researchers had further categorized acts like solving CAPTCHAs, reporting spam, and accepting or rejecting autocomplete suggestions as "unwitting labour"—performed without any awareness of contributing to AI training (Source: AI & Society, 2023. https://link.springer.com/article/10.1007/s00146-023-01692-3 ).

There is an instructive court case. In Rojas-Lozano v. Google, the plaintiff argued that solving reCAPTCHAs constituted uncompensated labor for Google. The court acknowledged that Google benefited, but dismissed the claim, reasoning that ordinary users would not expect payment and that no concrete loss could be established. In other words: even where one can say, economically and ethically, that value was created, current law provides no way to claim compensation. The institutional vacuum was confirmed in court.

The shared concern of all these thinkers can be put in one sentence: "Users are creating value, but they have no means of knowing the record, the purpose of use, the method of evaluation, or the amount of distribution."

Chapter 3: Policy Started Moving—Then Stopped

The ideas caught fire in politics.

In February 2019, California Governor Gavin Newsom declared in his State of the State address: "California's consumers should also be able to share in the wealth that is created from their data." He directed his administration to develop a "Data Dividend" proposal for Californians (Source: State of the State Address, 2019/2/12. https://www.gov.ca.gov/2019/02/12/state-of-the-state-address/ ).

Yet as of September 2026, no primary source shows that cash dividends to residents were ever implemented. California policy subsequently shifted its center of gravity from "the right to share in profits from data" to "the right to have data deleted" (the 2023 Delete Act and the DROP system effective 2026) (Source: California Privacy Protection Agency. https://cppa.ca.gov/regulations/drop.html ).

In New York, bill S3702—introduced in 2025 and still active—would tax companies deriving revenue from New Yorkers' data at 5% and distribute the proceeds to residents through a New York Data Fund (Source: NY Senate. https://www.nysenate.gov/legislation/bills/2025/S3702 ). At the federal level, the 2019 DASHBOARD Act would have required large platforms to measure and disclose the value of user data (Source: Sen. Warner press release. https://www.warner.senate.gov/newsroom/press-releases/warner-hawley-introduce-bill-to-force-social-media-companies-to-disclose-how-they-are-monetizing-user-data/ ).

International institutions moved as well. In 2021, the OECD adopted a Council Recommendation calling on member states to distribute the benefits of data sharing fairly and to build incentive systems in which contributing stakeholders are "recognized and rewarded"—a recommendation still in force (Source: OECD Legal Instruments. https://legalinstruments.oecd.org/api/print?ids=668&lang=en ). The World Bank's 2021 World Development Report proposed a "new social contract for data" built on Value, Equity, and Trust (Source: World Bank WDR2021. https://www.worldbank.org/en/publication/wdr2021 ).

So here is where we stand. The principle enjoys international support. Politicians gave speeches, bills were introduced, international institutions issued recommendations. But nowhere has anyone implemented a system that measures who contributed how much and returns value continuously. It is telling that California settled for deletion rights instead of dividends. Public policy could not solve the technically hard part: measuring and distributing contribution.

This is precisely why implementation must come from the private sector, from the technology side.

Chapter 4: The Value and the Losses, in Numbers

Let me anchor the economic scale of this future using only verifiable figures.

The size of the data economy. The European Commission estimates the EU27 data economy at €325 billion in 2019, or 2.6% of GDP; Japan's at 1.2% of GDP in the same year (Source: European Commission. https://policy.trade.ec.europa.eu/news/landmark-eu-japan-data-deal-one-step-closer-ratification-2023-12-01_en ). The OECD estimates the social and economic benefits of data access and sharing at 1–2.5% of GDP when private data is included, up to 4% in some studies (Source: OECD. https://www.oecd.org/en/publications/enhancing-access-to-and-sharing-of-data_276aaca8-en.html ). For an economy like Japan's, that means benefits on the order of trillions of yen, depending on how the circulation and reuse of data is designed.

Losses from fraud. Juniper Research estimated losses to digital advertising fraud at $68 billion in 2022 (Source: https://www.juniperresearch.com/press/digital-advertising-spend-lost-to-fraud-68-billion/ ). Online payment fraud is projected to reach $91 billion in 2028 alone (Source: https://www.juniperresearch.com/press/losses-online-payment-fraud-exceed-362-billion/ ). Thales (formerly Imperva) reports that a majority of internet traffic is now automated, with malicious bots alone accounting for roughly 40% of all traffic (Source: https://cpl.thalesgroup.com/about-us/newsroom/2025-imperva-bad-bot-report-ai-internet-traffic ). These figures cover different years and definitions and cannot be summed into a single "global fraud total." What can be said is this: the inability to verify that the other party is human is producing losses across multiple categories, each in the tens of billions of dollars.

The administrative impact of citizen participation. In the UK, FixMyStreet—which lets residents report potholes and broken streetlights from their phones—reduced telephone reports by 13% and email reports by 40% in one adopting council, where handling costs per report were £5.88 by phone and £7.81 by email (Source: mySociety. https://www.mysociety.org/2019/06/28/measuring-the-savings-brought-by-fixmystreet-pro/ ). Colorado optimized de-icing material use with vehicle-mounted road sensors, saving $180,000 in materials over three snowstorms, with projected savings of over $1 million per year statewide (Source: FHWA. https://www.fhwa.dot.gov/innovation/everydaycounts/edcnews/20190711.cfm ). Meanwhile, the U.S. GAO reports that about two-thirds of the 955 federal buildings under asbestos management failed to meet the five-year inspection policy—partly due to staffing shortages (Source: GAO. https://www.gao.gov/products/gao-24-106324 ). The things requiring inspection are growing; the people available to inspect them are shrinking. And it has already been demonstrated that citizens' everyday actions can fill that gap.

Chapter 5: Why reCAPTCHA and Waze, As They Are, Are Not Enough

The critiques of existing models are academically well organized. Four points stand out.

First, contributors cannot recognize their own contribution. The reCAPTCHA screen never shows which project you helped, or by how much. Second, the destination of the value is opaque. There is no way to check what your answers were used for, how many times they were reused, or how much cost the company saved. That the 2025 independent study had to estimate cumulative hours and value from the outside is itself proof that no official ledger of contribution exists. Third, service benefits and contribution compensation get conflated. "You use it for free, so no compensation is owed" mixes two different things: payment for using a service, and payment for the secondary use of your data. Fourth, the law offers no remedy. As the Rojas-Lozano ruling showed, there is a deep gap between "value was created" and "compensation can be claimed" under current law.

In sum, what existing models lack is an institution for the attribution, evaluation, visibility, and distribution of contribution. The creation of value is already happening. What is missing is the institution.

Chapter 6: What a Public Ledger Solves—and What It Does Not

So how do we build that institution? Only now does blockchain enter the discussion. The order matters: the problem comes first, not the technology.

In the classic analysis by Wüst and Gervais of ETH Zurich, a blockchain is compelling only when multiple parties write data, cannot fully trust one another, cannot appoint a single trusted administrator, and require external verifiability (Source: Wüst & Gervais, "Do you need a Blockchain?", 2018. https://eprint.iacr.org/2017/375 ). The recording and distribution of contributions fits these conditions exactly. The lesson of reCAPTCHA is that as long as contributions are recorded inside one company's private books, contributors can never verify them.

What a public ledger can guarantee: that a contribution record was registered by a certain time; that it has not been altered since; which version of the evaluation rules was applied; and where, and how much, was distributed. What it cannot guarantee is equally clear: whether a photo really depicts the specified equipment, whether GPS was spoofed, whether one person created multiple accounts. The ledger does not guarantee that the reality before registration was true. Moreover, putting personal data directly on a public ledger raises concerns—flagged by the European Parliament—about conflicts with the GDPR's right to erasure (Source: European Parliament resolution. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX%3A52018IP0528 ).

Hence the most realistic conclusion this field of research has reached is a hybrid architecture:

1. Personal and raw data stay in private domains, with access controls and deletion rights
2. The public ledger records only proofs of contribution (contribution ID, timestamp, data hash, verification result, evaluation-rule version, distribution amount)
3. The truthfulness of inputs—is this really a human? is there spoofing?—is verified by separate mechanisms: device attestation, duplicate detection, Sybil resistance, anomaly detection

Chapter 7: The Design Philosophy of DEP Protocol

If you have read this far, I can explain very briefly why our design takes the shape it does.

DEP (DEAPcoin) is a crypto asset issued by DEA. With the mainnet launch at 18:00 JST on July 14, 2026, it operates as the native token of DEP Protocol, a Layer 1 for human activity verification built on Avalanche. DEP Protocol is infrastructure that verifies and certifies human activity and generates human-verification scores for the AI, advertising, and robotics sectors.

This design matches the hybrid architecture of the previous chapter.

• No biometric or personal data is collected. The natural variability unique to human operation is judged inside the smartphone, and the underlying data never leaves the device. What is recorded on the ledger is only a human-verification score that identifies no individual, plus the records of verification and distribution
• Verifying the truthfulness of inputs is our core business. Against the limitation that public ledgers cannot guarantee input truth, DEP Protocol places the verification that a real human performed the action before the record
• Rewards are funded by real demand. Verification fees paid by those who need verification—AI companies checking the provenance of training data, advertisers confirming ads were seen by humans, enterprises and governments proving human involvement in approvals—are distributed to the validators who operate the infrastructure. This is not a reward design dependent on newly minted tokens. Many of the models once called "x-to-earn" failed to sustain themselves precisely because their rewards were funded by token inflation rather than external demand. We designed on top of that lesson

And we should not forget why this mechanism became necessary. More than half of internet traffic is no longer human. The EU has enacted a law mandating human oversight for high-risk AI applications, with enforcement for those uses beginning in December 2027 (Source: Regulation (EU) 2024/1689; timeline amended by Regulation (EU) 2026/1744). The question "was this really done by a human?" has become one that regulation and business alike are willing to pay to have answered.

Chapter 8: Every App Becomes a Doorway to Contribution

Here is the future we see.

When an app or web service integrates the DEP Protocol SDK, the operations of its users generate human-verification data that identifies no individual. This is useful not only for "defense"—bot exclusion, duplicate-application prevention. The verification data itself becomes a supply that answers the real demand of AI companies and advertisers. In other words:

The everyday actions of people simply using an app become, outside that app's original purpose, a supply of the "proof of humanity" that society needs—and that contribution is recorded, evaluated, and compensated, with value flowing back to the app and its users.

What reCAPTCHA proved is that this structure works technically and generates enormous value. All that was missing was the institution for attribution, evaluation, and distribution. What Lanier called data dignity, what the OECD recommended as systems where contributors are "recognized and rewarded," what California dreamed of but could not implement—we are implementing it, through the combination of a verifiable ledger, demand-based distribution, and human verification that identifies no individual.

Reality is steadily moving in this direction. In August 2026, the humanoid robotics company Figure AI announced Index, an app where ordinary people submit videos of everyday actions and receive compensation: over 260,000 downloads across 100+ countries, more than 16 million videos submitted, over $15 million paid to contributors, and a commitment to spend more than $1 billion on data and compute over the next twelve months (Source: Figure AI. https://www.figure.ai/news ). Companies that pay for human behavioral data now exist in the real world. And their pipeline explicitly includes a fraud-review stage—because wherever rewards arise, so does the question: "is this really a human?"

What we have built up in Japan—citizen-participation infrastructure inspection through PicTrée, community missions through our regional discovery app—are early implementations of this future. People walk their towns for fun, take photos, answer quizzes. That daily life helps maintain infrastructure, helps train AI, and improves their communities. The contribution is visible to them, and value returns to them.

Epilogue: The World We Want to Build

Let me line up what this essay has established.

The phenomenon of everyday actions generating social value is already occurring at the scale of billions to hundreds of billions of events (reCAPTCHA, Waze, BOINC). Contributions can be quantified—in books transcribed, minutes of travel saved, material costs reduced, inspection points covered. Thinkers named this invisible labor more than twenty years ago; economists proposed institutionalizing it; politicians gave speeches; international bodies issued recommendations. It was never implemented because the technically hard part remained: attribution, evaluation, and distribution of contribution.

That part is now solvable. The architecture that records only proofs of contribution while protecting personal data has been identified by research as the most realistic approach—and the technologies for human verification and demand-based distribution are now in place.

Ordinary actions by ordinary people, making society better. Contributions that are visible to the contributor, fairly evaluated, and rewarded. Not inside someone's private books, but in a form anyone can verify.

That is the world DEP Protocol is building.

References

1. von Ahn, L. et al. "reCAPTCHA: Human-Based Character Recognition via Web Security Measures." Science, 2008. https://www.cs.cmu.edu/~biglou/reCAPTCHA_Science.pdf
2. NDSS Symposium USEC 2025 (independent estimates of cumulative reCAPTCHA labor). https://www.ndss-symposium.org/wp-content/uploads/usec25-21.pdf
3. Google Security Blog. "Street View and reCAPTCHA technology." 2014. https://security.googleblog.com/2014/04/street-view-and-recaptcha-technology.html
4. Waze for Cities / Joinville Case Study. https://www.waze.com/wazeforcities/casestudies/using-data-for-roadway-upgrades/
5. City of Boston. "Street Bump." https://www.boston.gov/departments/new-urban-mechanics/street-bump
6. mySociety. "Measuring the savings brought by FixMyStreet Pro." 2019. https://www.mysociety.org/2019/06/28/measuring-the-savings-brought-by-fixmystreet-pro/
7. FHWA. "Colorado DOT Mobile Road Weather Sensors." 2019. https://www.fhwa.dot.gov/innovation/everydaycounts/edcnews/20190711.cfm
8. GAO-24-106324 (federal building asbestos inspection shortfalls). https://www.gao.gov/products/gao-24-106324
9. Terranova, T. "Free Labor: Producing Culture for the Digital Economy." Social Text 18(2), 2000. https://web.mit.edu/schock/www/docs/18.2terranova.pdf
10. Lanier, J. "Who Owns the Future?" 2013. https://www.simonandschuster.com/books/Who-Owns-the-Future/Jaron-Lanier/9781451654974
11. Arrieta-Ibarra, I., Goff, L., Jiménez-Hernández, D., Lanier, J., Weyl, E.G. "Should We Treat Data as Labor? Moving beyond 'Free'." AEA Papers and Proceedings 108, 2018. https://www.aeaweb.org/articles?id=10.1257%2Fpandp.20181003
12. "The unwitting labourer: extracting humanness in AI training." AI & Society, 2023. https://link.springer.com/article/10.1007/s00146-023-01692-3
13. California Governor. State of the State Address, 2019. https://www.gov.ca.gov/2019/02/12/state-of-the-state-address/
14. California Privacy Protection Agency. DROP regulations. https://cppa.ca.gov/regulations/drop.html
15. NY Senate Bill S3702. https://www.nysenate.gov/legislation/bills/2025/S3702
16. Sen. Warner. DASHBOARD Act press release, 2019. https://www.warner.senate.gov/newsroom/press-releases/warner-hawley-introduce-bill-to-force-social-media-companies-to-disclose-how-they-are-monetizing-user-data/
17. OECD. Recommendation on Enhancing Access to and Sharing of Data, 2021. https://legalinstruments.oecd.org/api/print?ids=668&lang=en
18. OECD. "Enhancing Access to and Sharing of Data." https://www.oecd.org/en/publications/enhancing-access-to-and-sharing-of-data_276aaca8-en.html
19. World Bank. World Development Report 2021: Data for Better Lives. https://www.worldbank.org/en/publication/wdr2021
20. European Commission. EU-Japan data flows (data economy figures). https://policy.trade.ec.europa.eu/news/landmark-eu-japan-data-deal-one-step-closer-ratification-2023-12-01_en
21. Juniper Research. Digital Advertising Fraud, 2022. https://www.juniperresearch.com/press/digital-advertising-spend-lost-to-fraud-68-billion/
22. Juniper Research. Online Payment Fraud, 2023. https://www.juniperresearch.com/press/losses-online-payment-fraud-exceed-362-billion/
23. Thales. 2025 Imperva Bad Bot Report. https://cpl.thalesgroup.com/about-us/newsroom/2025-imperva-bad-bot-report-ai-internet-traffic
24. Wüst, K., Gervais, A. "Do you need a Blockchain?" 2018. https://eprint.iacr.org/2017/375
25. European Parliament resolution on distributed ledger technologies (right to erasure). https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX%3A52018IP0528
26. Regulation (EU) 2024/1689 (EU AI Act); Regulation (EU) 2026/1744 (timeline amendment)
27. Figure AI. Index announcement. https://www.figure.ai/news
28. DEP Protocol official site. https://dep-protocol.com/

Note: Among the estimates cited, the NDSS figures ($6.1B, etc.) are external researchers' approximations, and the Waze Joinville economic impact is Waze's own published case-study figure. Neither is an independently audited value.