GPT-3 and AI Freedom
August 2021
Much has been said about the Artificial Intelligence language model GPT-3 (Generative Pre-trained Transformer 3) by OpenAI recently. Its capabilities are impressive, and it promises technological and social disruption that has the potential to be both benevolent and malicious.
- TechTalks – What OpenAI and GitHub’s “AI pair programmer” means for the software industry ”Neural networks are very good at finding and suggesting patterns... In this light, it makes sense to use GPT-3 … to help programmers find solutions in the very large corpus of publicly available source code in GitHub.”
- Analytics India Magazine – Alchemists And Altruists: Who Should Hold The Key To The Powers Of AI? ”Malicious uses of models can be difficult to anticipate because they can be repurposed in a very different environment or for a different purpose than what the researchers intended.”
Left solely in OpenAI’s hands, GPT-3 would remain a black-box proprietary technology with all the restrictions and limitations associated with using software as a service provided by a third party. The ability to freely explore the technology’s capabilities and develop countermeasures for its negative impacts would remain restricted by an organisation making profit-driven decisions about access and control.
To counter this, EleutherAI, a collective of volunteer researchers, engineers, and developers, are working on an open source competitor to GPT-3. The advantages include being able to control the source code and software yourself, train and configure the model on your own domain-specific data, and integrate the technology directly into your own software and business tools without relying on a middle-man.
Further reading:
- The Next Web explains – Understanding Transformers, the machine learning model behind GPT-3 ”How this novel neural network architecture changes the way we analyze complex data types.”
- EleutherAI explains – Why Release a Large Language Model? “We believe the creation and open source release of a large language model is a net good to AI safety. We explain why.”
- VentureBeat – With post-pandemic AI, we’ve now stepped into the Age of Acceleration ”Google is using AI to design chips that can be used to create even more sophisticated AI systems, further speeding-up the already exponential performance gains through a virtuous cycle of innovation.”
Ashley Flynn – Games and Simulations Software Engineer Portfolio and contact – https://ajflynn.io/