The Health Transformation
institute for health strategy, digital health
and continuous transformation
Joaquim Cardoso MSc
Chief Research and Strategy Officer (CRSO)
May 16, 2023
ONE PAGE SUMMARY
Tech companies like Google seek processing power in handsets to reduce computing costs and improve speed of AI chatbots
Based on the article “The race to bring generative AI to mobile devices”, published by the Financial Times, on May 16, 2023.
The race is on to bring generative AI technology, similar to ChatGPT, to mobile devices.
Tech companies like Google are seeking ways to run generative AI on smartphones rather than relying on cloud-based servers to reduce computing costs and improve the speed of AI chatbots.
The ability to run generative AI on mobile handsets would make services like chatbots more affordable for companies and enable transformative applications using generative AI.
Recent advances in smaller language models have made these capabilities more accessible and less computationally demanding.
Open-source models like LLaMa have demonstrated surprising capabilities with smaller sizes, making them adaptable for different computing environments.
While the development of mobile applications using generative AI is still in the experimental stage, it is expected that the first wave of mobile models will focus on relatively rudimentary apps such as voice-controlled photo editing and simple question-answering.
Mobile applications are likely to rely more on speech and images rather than text-heavy applications.
Apple is expected to take a cautious approach to embedding generative AI into its iPhone operating system …
… due to well-known flaws in large models, while focusing on providing tools for app developers to experiment with the technology.
