about the reality gap between businesses talking about artificial intelligence and those that are actually using it, an investment analyst from Oppenheimer said, “There’s not a single call I sit on these days where AI is not mentioned in some form." The number of businesses that actually implement this technology, however, is"conservative." Additionally, a,"While companies race to incorporate AI internally and externally, consumer use of genAI is still in its infancy.
During these early stages of the AI revolution, a great deal of attention has been devoted to large language models . By most measures, from simple Google searches to refined SEO measurements, LLMs dwarf small language models . But as I predicted in a, concerns about the expense and compute power required for LLMs are compelling practical executives to seek alternatives. In fact, the development of SLMs is progressing at a faster rate than I expected.
Regardless of the adoption rate, the reality is that to progress toward artificial general intelligence systems—LLMs and SLMs—companies need functionality that hasn't been generally available in the past. The demand for more data in more places leads to greater complexity and skyrocketing costs. Meanwhile, performance and reliability often plummet.Vector databases have received an enormous amount of publicity in recent months.