Now more than ever, a governance framework is needed

Free updates are notified at any time

The author is Founding Co-Director of Stanford University’s Human-AI (HAI) and CEO and Co-Founder of World Labs

Artificial intelligence is moving forward at an astonishing speed. Computing model days can now be done in minutes, and despite the sharp rise in training costs, they will soon drop as developers learn less. I've said it before, and I'll say it again – the future of AI is now.

This is not surprising to anyone in the field. Computer scientists have been working hard. The company has been innovating for many years. Surprisingly, eyebrow burrs – it seems that there is a lack of an overall framework for AI governance. Yes, AI is developing rapidly – it is necessary to ensure it benefits all human beings.

As a technician and educator, I strongly believe that everyone in the global AI ecosystem is responsible for advancing technology and a people-centered approach. It's a difficult task and deserves a structured set of guidelines. In preparation for the AI Action Summit in Paris next week, I have put forward three basic principles for the future of AI decision-making.

First, use science, not science fiction. The basis of scientific work is the principle of dependence on empirical data and rigorous research. The same approach should apply to AI governance. While futuristic scenarios capture our imaginations (whether it is a utopia or apocalypse), effective decisions require a clear view of current reality.

We have made significant progress in areas such as image recognition and natural language processing. Chatbots and co-pilot software assistance programs are changing jobs in exciting ways, but they are applying advanced data learning and pattern generation. They are not intellectual forms of intention, free will, or consciousness. It is crucial to understand this, save us from distracted scenarios and keep us focused on important challenges.

Given the complexity of AI, it is not always easy to focus on our reality. To bridge the gap between scientific advancements and real-world applications, we need tools that will share accurate, latest information about their capabilities. Established institutions, such as the National Institute of Standards and Technology, can illuminate the reality of AI, resulting in exact, feasible policies based on technical reality.

Secondly, be pragmatic, not ideology. Despite the rapid development of AI, it is still in its infancy and has made the most contribution. In this case, policies that can and cannot be formulated must be pragmatically to minimize unexpected consequences while inspiring innovation.

For example, use of AI for more accurate diagnosis of diseases. This has the potential to rapidly democratize access to high-quality care. But without the right guidance, it could also exacerbate biases that exist in today’s healthcare systems.

Developing AI is not easy. It is possible to develop a model with the best intentions and abuse it later. Therefore, the best governance policies will aim to reduce this risk tactically while rewarding responsible implementation. Policy makers must develop practical responsibility policies to prevent intentional abuse without unfairly punishing goodwill efforts.

Finally, empower the AI ecosystem. This technology can inspire students, help us take care of an aging population, and provide innovative solutions for clean energy and achieve the best innovation through collaboration. Therefore, it is more important for decision makers to empower the entire AI ecosystem, including the open source community and academia.

Open access to AI models and computing tools is crucial to progress. Limiting it will create obstacles and slow innovation, especially for researchers with fewer academic institutions and resources than the private sector. Of course, the consequences of this limitation go far beyond the academic community. If today's computer science majors are unable to use the best model for research, then they won't understand these complex systems when they enter the private sector or decide to find their own company – a serious gap.

The AI revolution is here – I'm very excited. It is possible that we can significantly improve our human condition in an AI-driven world, but to achieve this reality we need experience, collaborative and deeply rooted in governance of people-centered values.