Despite the current hype around generative AI, some industry insiders believe we are nearing the limits of the current gen AI boom. Venture capitalists Paul Kedrosky and Eric Norlin of SK Ventures argue that substantial technological improvements are needed to achieve better than "so-so automation" and limited productivity growth. They predict the next wave of AI will include new models, more open source, and notably "ubiquitous/cheap GPUs". This could challenge Nvidia's dominance but would benefit those needing the technology. Despite the existing limitations of the current generation of models and applications, the future for gen AI appears bright, driven by a generational shortage of workers across the economy that will continue to drive the need for greater automation.
Key takeaways:
- Generative AI is gaining wider adoption in businesses, with Walmart rolling out a gen AI app to 50,000 non-store employees. This is driving demand for graphical processing units (GPUs) needed to train powerful deep learning models.
- Nvidia, the dominant supplier of GPUs, is experiencing a surge in demand for their products. Google and Nvidia have announced a partnership to give Google’s cloud customers greater access to technology powered by Nvidia’s GPUs.
- However, not all software benefits from the parallel instruction execution of GPUs. Many existing business applications are optimized to run on CPUs and would see little benefit from GPUs.
- Despite the current limitations of gen AI, the future for the technology appears bright. McKinsey believes that generative AI is poised to unleash the next wave of productivity, potentially raising global GDP by 7%.