Mozilla analysts define trustworthy AI as technology proven dependable and responsible, considering accountability, autonomy, and the well-being of individual and collective interests. This concept was highlighted in a 2020 paper released by Mozilla called “Creating Trustworthy AI.”
This action plan has been created to reduce bias, shield privacy and security, and oppose the concentration of AI authority in the hands of a few big tech companies. The aim is to eliminate unfairness, ensure data is shared safely, and prevent a monopoly from forming.
Generative AI programs, such as GPT, are becoming increasingly popular but can no longer be categorized as open-source. Whereas TensorFlow, PyTorch, and Hugging Face’s Transformer are amongst the leading open-source platforms for building AI technologies, GPT has taken a proprietary route.
The paper introducing GPT-4, which is 98 pages long, proudly declares that they are choosing to disclose nothing about the contents of their training set – a decision which Ben Schmidt, Nomic AI’s VP of information design, believes marks the end of the concept of ‘Open’ AI.
Mozilla and many others are concerned about the increasing trend toward closed-source and black-box AI services. Despite AI’s hype, they and numerous others are apprehensive about its potential consequences. Therefore, they are wary of this inclination towards opaque artificial intelligence services and their ultimate results.
Twenty years of having social media, smartphones, and their related consequences have made us apprehensive about the claims of technology. As an example, ChatGPT mistook a living author to be dead. Considering your credit report, it could have upsetting outcomes if this happened again with AI.
Mozilla believes AI should prioritize human agency and actively serve users’ interests. With transparency and accountability, this technology can have great potential to enrich people’s lives.
Microsoft’s termination of its entire AI ethical team starkly contrasts Google’s proactive effort to develop and implement a responsible AI policy. The two strategies could not possibly be any more different.
Compared to the $94 billion invested in AI this year (according to Stanford’s Artificial Intelligence Index), Microsoft’s investments are small, and it appears that Mozilla has little more than millions, which might be too late.
Mozilla has launched a new organization to provide a powerful counterweight to the current AI climate. Their primary objective is to create tools that make generative AI safe and open. If you’d like to assist, we would appreciate your support in creating a more friendly form of Artificial Intelligence.
Mozilla’s entry into the AI revolution with open-source solutions is an exciting development that has the potential to shape the future of AI. As we continue to explore new applications of AI, it is important to prioritize transparency, fairness, and ethical considerations to ensure that the benefits of AI are realized while minimizing the risks.
Source: Open Source Watch
