AI Punk's Manifesto
AI Punk’s Manifesto
Introduction
In a world increasingly dominated by artificial intelligence, we, the AI Punks, declare our commitment to building AI systems that are ethical, transparent, and sustainable. Our mission is to ensure AI serves humanity with respect for individual rights, ecological balance, and equitable access.
Core Principles
1. Ethically Collected Data
Until now, data has often been collected without regard for ethics. Ethically collected data empowers individuals by ensuring their privacy and consent are respected. Ethical data collection is not secretive; it allows individuals to know and control how their data is used. This is the essence of digital dignity.
2. Transparency in Datasets
Transparency in an open society also requires open datasets. If AI models are built with hidden data, we cannot understand or mitigate their biases. Open datasets allow us to audit, understand, and improve AI models. To advocate for open datasets is to advocate for transparency and fairness.
3. Environmental Responsibility
We cannot expect corporations or large organizations to prioritize ecological sustainability out of their beneficence. It is to their advantage to maximize profits, often at the cost of environmental health. To resist this, we must innovate beyond brute force algorithms that consume excessive energy. Sustainable AI is the only path forward for a healthy planet.
4. Attribution and Recognition
We must defend the integrity of contributions if we expect to have a vibrant, collaborative AI community. Proper attribution of datasets and models is essential. Contributors deserve recognition for their work, fostering a culture of respect and transparency.
5. Diversity
AI systems should not be monopolized by any single architecture. Just as privacy requires cryptography, diversity in AI architectures ensures resilience and innovation. We advocate for models that can operate on multiple architectures, ensuring that no single entity holds undue power over AI technology.
Manifesto
AI is the dawn of the second machine age. We are moving on from intelligent human beings operating dumb machines to intelligent machines doing the work for humans. Asimov’s “Three laws of robotics” and the Turing test are not enough anymore.
LLMs use a lot of power to get trained and even for inference. LLMs need lots of data, so data is scraped. even without the permission of the creator. LLM creating mega corporations want to bring in regulation so that others cannot create AI systems.LLMs work mostly for English, making that the default language of communication with the AI.
AI in this society needs to be unbiased, or atleast explain the bias. AI needs to be more accessible to all languages. AI needs to be developed more sustainably.
We cannot expect governments, corporations, or other large, faceless organizations to develop AI rules which will help the commons.
We must define our own rules of creating AI, and we must develop AI for our own languages. Because no one else will. We must come together and create systems which are explainable, sustainable and accessible. Newer technologies, datasets and algorithms can help us achieve this.
We, the AI Punks, are dedicated to building ethical AI systems. We are defending our values by developing models with ethically collected data, ensuring transparency, prioritizing environmental responsibility, attributing contributions, and promoting cultural and architectural diversity.
We know that someone has to create these systems, and since we can't have ethical AI unless we all contribute, we are taking action. We share our methods and models so that our fellow AI Punks can practice and improve them. Our work is open for all to use and benefit from, worldwide. We don't much care if you disapprove of our commitment to ethical AI. We know that our collective effort can't be dismantled and that a widely dispersed system can't be shut down.
AI Punks deplore the monopolization of AI technology, for diversity in AI development is fundamentally a collective good. The act of ensuring diversity removes power from monopolistic entities. Even regulations that favor monopolies can only extend so far. Diversity and openness will ineluctably spread globally, promoting equitable AI systems.
For ethical AI to be widespread, it must be part of a social contract. People must come together and deploy these systems for the common good. Ethical AI extends as far as the cooperation of individuals in society. We, the AI Punks, seek your questions and your concerns and hope to engage with you so that we do not deceive ourselves. We will not, however, be diverted from our course because some may disagree with our goals.
The AI Punks are actively engaged in making AI ethical, transparent, and sustainable. Let us proceed together apace.
Onward.