d/acc: A Review and Outlook One Year Later

Written by: Vitalik Buterin

Foreword

About a year ago, I wrote an article on technological optimism, outlining the enthusiasm for technology and its immense benefits, while also expressing a cautious attitude towards certain issues, particularly concerning the risks that superintelligent AI may bring.

My core point in that article is to advocate for a decentralized, democratic, and differentiated defensive acceleration concept. This means accelerating technological development, but focusing on technologies that enhance defensive capabilities rather than destructive ones, and striving to decentralize power to avoid a situation where a few elites represent everyone in determining right and wrong. The defensive model should resemble a democratic Switzerland, rather than a castle under medieval feudalism.

Over the past year, these concepts and ideas have significantly developed and matured. I shared these thoughts on the "80,000 Hours" platform and received a lot of feedback, most of which was positive, though there were also some critical voices.

This work has continued to advance and has achieved tangible results: progress has been made in areas such as verifiable open-source vaccines, healthy indoor air, "community notes," prediction markets, the application of zero-knowledge proofs in government identification and social media, and open-source medical imaging tools.

Last autumn, we held our first major d/acc event: "d/acc Discovery Day" ( d/aDDy ), which brought together speakers from various pillar areas of d/acc, allowing those who have long been dedicated to these technologies to better understand each other's work, and also making outsiders aware of this broader vision: the values that drive the development of Ethereum and cryptocurrency can extend to a wider world.

The Connotation and Denotation of d/acc

The core idea of d/acc is: decentralized, democratic, and differentiated defensive acceleration. It builds technologies that facilitate a balance between offense and defense leaning towards defense, and does not rely on transferring more power to central authority during the implementation process. These two aspects are intrinsically linked: any decentralized, democratic, or free political structure tends to thrive when defense is easy to implement, while it faces severe challenges when defense is difficult.

One way to understand and implement the importance of decentralization, defensiveness, and acceleration is to compare it with the ideas produced by abandoning any one of these three aspects:

1. Decentralization accelerates, but neglects the "differentiated defense" part.
2. Differentiated defense acceleration, but neglecting "decentralization and democracy"
3. Decentralized defense, but exclusion of acceleration

Through d/acc, we are committed to achieving the following goals:

• Upholding principles in the increasingly tribalized trend of today's world, building specific things to make the world safer and better.
• Recognizing that exponential technological advancements mean the world will become extremely peculiar, humanity's "footprint" in the universe will inevitably continue to grow. Our ability to protect fragile things must continuously improve, and the only way forward is to move boldly ahead.
• Build technologies that can genuinely protect us, rather than relying on the assumption of "good people ( or good AIs ) controlling everything." We achieve this goal by creating tools that are naturally more effective when used for building and protecting than when used for destruction.

Another perspective on d/acc is to return to the framework of the Pirate Party movement in Europe at the end of the 2000s: empowerment. Our goal is to build a world that can preserve human agency, achieving both negative liberty and positive liberty. This resonates with a classical liberal tradition that has lasted for centuries. Given the technological landscape of the 21st century, we can view d/acc as a means to achieve these same goals.

Challenges: AI Safety, Urgent Timelines, and Regulatory Dilemmas

Last year, the most compelling opposing viewpoint to my article came from the AI safety community. Their argument is: it seems we may have only 3 years to develop to general AI, and another 3 years to develop to superintelligence. Therefore, if we do not want the world to fall into destruction or an irreversible predicament, we cannot simply accelerate the development of beneficial technology; we must also slow down the development of harmful technology, which means that strong regulatory measures may be needed that could anger the powerful.

If we are in the least ideal world, where the AI risks are extremely high and the timeline may be as short as 5 years, I would support the following regulatory measures:

1. Responsibility
2. Global "soft pause" button on industrial-scale hardware

These strategies have loopholes and are only temporary measures. We need more stable measures to buy time. Many d/acc technologies are relevant here and can help us cope with the potential ways AI could take over the world:

• Network defense against AI invading our computers
• Biological defense against AI-manufactured super plagues
• Information defense should respond to AI persuading us.

Responsibility rules are a regulatory approach that naturally aligns with the d/acc philosophy, as they can effectively incentivize the adoption of these defensive measures worldwide and take them seriously.

The Role of Cryptocurrency in d/acc

There are some important connections between cryptocurrency and d/acc:

• d/acc is the extension of the fundamental value of cryptocurrency to other technological fields.
• The cryptocurrency community is the natural early user and incubator of d/acc technology.
• Many cryptocurrency technologies can be applied to the d/acc thematic field.
• There are win-win cooperation opportunities in adjacent technologies.

d/acc and Public Goods Funding

Powerful decentralized public goods funding is crucial for the d/acc vision, as one of the key goals of d/acc is to minimize central control points, which itself hinders many traditional business models. Therefore, we must do the hard thing, which is to figure out how to fund public goods in a way that addresses the criticisms.

Deep funding is my latest effort to address this issue. It innovates by using dependency graphs and artificial intelligence as refined human judgment. The fundamental goal is to create a social tool that can fund public goods with accuracy, fairness, and open access levels that are at least close to those of market-funding private goods.

Future

The coming decades bring significant challenges:

1. A powerful wave of new technologies, especially strong AI, is rapidly approaching, accompanied by important pitfalls that we need to avoid.
2. The world is becoming increasingly uncooperative.

However, each of these challenges has a glimmer of hope:

1. We now have very powerful tools to complete the remaining work faster.
2. Many of the principles we cherish can be reclaimed by a broad alliance that welcomes anyone in the world to join.

The acquisition of tools means that we can adapt and improve our biological characteristics and environment, and the "defense" aspect in d/acc means that we can do this without infringing on the freedoms of others doing the same. The principle of freedom pluralism means that we can have great diversity in how we achieve this, and our commitment to common human goals means that it should be realized.

We humans are still the brightest stars. Building a brighter 21st century while protecting human survival, freedom, and agency in the process of reaching for the stars is a challenging task. But I believe we can rise to the occasion.