Evidence that GPTs cannot be smarter than their training data

In today’s email:

Google researchers have submitted a pre-print paper to ArXiv which casts doubt on the ability of transformer technology, which underpins large language models (LLMs) like ChatGPT, to generalize beyond their training data. The paper points out that while these AI models are effective at tasks they are trained on, they struggle with out-of-domain tasks and simple extrapolation, raising questions about the current feasibility of achieving artificial general intelligence (AGI). The research suggests that while transformers represent an advanced technology, the complexity and opacity of neural networks combined with the vast amounts of training data may have led to overestimated capabilities. This realization is making many in the field acknowledge the limitations of LLMs. Experts and industry leaders suggest that although transformers may not be the final solution to AGI, they are still a powerful tool that can be directed and aligned for practical uses.

Study artificial intelligence and gain the knowledge to support its integration into your organization. If you’re looking to gain a competitive edge in today’s business world, then this artificial intelligence online course may be the perfect option for you.

• Key AI management and leadership insights to support informed, strategic decision-making.

• A practical grounding in AI and its business applications, helping you to transform your organization into a future-forward business.

• A road map for the strategic implementation of AI technologies in a business context.

Superpower ChatGPT 5.0.0 has been released. 🎉

Unclassified 🌀 

Hope you enjoyed today’s newsletter

⚡️ Join over 200,000 people using the Superpower ChatGPT extension on Chrome and Firefox.