Businesses Should Know Exactly How Their AI Models Are Trained
The Sora Incident: Triggering a Much-Needed Conversation on Transparency and Ethical AI in Business
The recent incident involving OpenAI's CTO, Mira Murati, not knowing the specific data used to train their text-to-video model, Sora AI, is a reminder of the importance of transparency and understanding in AI development. This lack of clarity from a leading AI firm expected to pioneer Artificial General Intelligence (AGI) is concerning, to say the least.
However, this incident also presents an opportunity to reflect on what AI should mean in the business world. At Naas, we're provoking this conversation by naming our main project, Artificial Business Intelligence (ABI).
ABI is more than just a play on words. It's a call to redefine what AI means in a business context. While AGI aims to create a machine with general intelligence across various tasks, ABI is designed to augment business processes. It's about leveraging AI to automate tasks, analyze data, and provide insights to help businesses thrive.
But ABI is not just about functionality. It's also about transparency, accountability, and practicality. Businesses should know exactly how their AI models are trained, what data they use, and how they make decisions; we advocate for open source. This level of transparency is crucial for building trust and ensuring that AI is used responsibly.
The incident with OpenAI underscores the importance of this approach. It's not enough to have powerful AI technologies. We must also understand them, control them, and use them ethically. This is what ABI stands for.
By naming our project ABI, we're making a statement. We're saying that AI in business should be intelligent, yes, but also ethical, transparent, and accountable. It's a provocation, a call to action, and a commitment to do the work to make AI work.
Join us in this conversation. Read our freshly updated documentation, and contact me to chat! Now that we are moving from alpha to beta, I will explain more about what we have learned over the past months.
Thanks for reading this rant and self-promotion :)
Jeremy
Looking forward to learning more about what you’re doing 😊 I have a long draft Substack article in progress on ethics of genAI in music, and this need for transparency is exactly one of our callouts!