In the long term, companies like D-Wave are exploring the use of quantum processing units (QPUs) in the training process and the application of quantum computing to sampling. French quantum computing startup Pasqal is also looking at using quantum computing to offload graph structured data sets commonly found in neural networks. However, this would require quantum systems to become significantly larger and faster. Bartlett's expertise in silicon photonics, a technology that could overcome bandwidth limits and scale machine learning performance, could also be of interest to OpenAI.
Key takeaways:
- OpenAI has hired Ben Bartlett, a former quantum systems architect at PsiQuantum, potentially signaling a move towards quantum computing to improve AI model efficiency.
- Quantum computing could drastically improve the efficiency of training large AI models, allowing them to derive more accurate answers from models with fewer parameters.
- Quantum algorithms can be used to optimize AI training datasets for specific requirements, resulting in leaner, more accurate models.
- OpenAI might also be interested in silicon photonics, a technology that Bartlett has expertise in, which could help overcome bandwidth limits and scale machine learning performance.