OpenAI charter puts safety, standards, and transparency first

OpenAI, the non-profit that promotes the development of artificial intelligence, has released a charter. In it, the organization outlines the core principles it believes should govern the development and management of artificial intelligence. The OpenAI charter represents an important step in initiating a broader discussion around the ethical considerations of artificial intelligence.

Revealed in a short blog post, the organization explains that the OpenAI charter is a summation of the development of its strategy over the last two years. Its mission remains central to the charter, however: ensuring that the development of artificial intelligence benefits all of humanity. 

What's inside the OpenAI charter?

The charter is then broken down into 4 other areas.

• Broadly-distributed benefits - OpenAI claims its primary duty is to humanity
• Long-term safety
• Technical leadership - OpenAI places itself at the cutting edge of the technology that will drive AI forward
• Cooperative orientation - working with policy-makers and institutions

Core concerns the OpenAI charter aims to address

A number of core concerns lie at the heart of the charter. One of the most prominent is what OpenAI see as the competitive race to create AGI "without time for adequate safety precautions". It's because of this that OpenAI seeks cooperation with "other research and policy institutions" - essentially ensuring that AI doesn't become a secretive corporate arms race. Clearly, for OpenAI, transparency will be key to creating artificial intelligence that is 'safe'.

OpenAI also claims it will publish its most recent AI research. But perhaps even more interestingly, the charter goes on to say that "we expect that safety and security concerns will reduce our traditional publishing in the future, while increasing the importance of sharing safety, policy, and standards research." There appears to be a tacit recognition of a tension between innovation of AI and the ethics around such innovations.

A question nevertheless remains over how easy it is for an organization to be at the cutting-edge of AI technology, while taking part in conversations around safety and ethics. As the last decade of technical development has proved, innovation and standards sometimes seem to be diametrically opposed, rather than in support of one another. This might be important in moving beyond that apparent opposition.

• 'If tech is building the future, let’s make that future inclusive and representative of all of society’ – An interview with Charlotte Jee
• What your organization needs to know about GDPR
• 20 lessons on bias in machine learning systems by Kate Crawford at NIPS 2017
• Mark Zuckerberg’s Congressional testimony: 5 things we learned
• The Cambridge Analytica scandal and ethics in data science