Andrew Ng on AI Progress: Far from Hitting a Wall

New financial motivations are helping to solve the supply chain problems for GPUs and other hardware. This means more AI deployments in the next few years. Even if AI stopped creating new tech today, many projects will still launch soon. There's even better news: new tech is always on the horizon. This will drive even more applications in the future.

Many validated ideas already exist that can drive real returns. These ideas haven't been deployed yet due to various reasons. With the supply chain for GPUs improving, more projects will get the green light. This is why many experts feel confident about the future of AI projects. Right now, the main bottleneck is hardware supply. As more GPUs get made, more projects will start.

MidJourney has recently released version 6.1. The images it produces are almost indistinguishable from photography. They have improved arms, legs, bodies, plants, and animals. The quality of the images is much better. They have also reduced pixel artifacts and enhanced textures and skin. One main improvement is in text accuracy. MidJourney version 6 had issues with text accuracy, so many users had to use different models like Ideogram or GPT.

MidJourney has dominated the AI generation space for a long time. Recently, a new competitor named Flux has entered the scene. Flux might change the game, but not for the reasons most people think. Flux's technology offers a new way to generate images that many find revolutionary. It has forced other companies to rethink their strategies and improve their models faster.

AI advancements are coming at a rapid pace. This includes better image generation and improved hardware supply chains. These changes mean more projects can start and succeed. With companies like MidJourney and Flux pushing the boundaries, the future looks bright for AI technology.