Rather than asserting that current LLMs are at their tail end, or that AI isnt good enough, it is much more instructive to check what are the bottlenecks or constraints to further progress, and what would help remove these bottlenecks.<p>They can largely be divided into 3 buckets<p>1) Compute constraint - Currently large companies using expensive nvidia chips do most of the heavylifting of training good models. Although chips will improve over time, and competition like Intel/AMD will bring down prices, this is a slow process. But what could be a faster breakthrough is training using distributed computing over millions of consumer GPUs. There are already efforts in that direction (eg. petals/swarm parallelism for finetuning/full training, but the eastern europe/russian guys developing them dont seem to have enough resources).<p>2) Data constraint - If you just rely on human generated text data, you will soon exhaust this resource (maybe GPT4 has already). But the Tinystories dataset generated from GPT4 shows if we can have SOTA models generate more data (and especially on niche topics that appear less frequently in human generated data), and have deterministic/AI filters to segregate the good and bad quality data thus generated, data quantity would not be an issue any longer. Also, multimodal data is expected (with the right model architectures) to be more efficient at training world grokking SOTA models than single modal data and here we have massive amounts of online video data to tap into.<p>3) Architectural knowledge constraint - This may be the most difficult of all, figuring out what is the next big scalable architecture after Transformers. Either we keep trying newer ideas (like the stanford hazy research group does), and hope something sticks, or we get SOTA models few years down the line to do this ideation part for us.