OpenAI have dominated the global headlines in November 2022, bashing with a new era of artificial intelligence that promises to change the world as we knew it. It captivated strungs of baffled humans not knowing what to make of their future. Yet it paved the way for a new tech race, one that is continuosly collecting speed and dominance. It introduced an unprecedented boost to individual productivity, no denial at all. But only for the individual.
Starting yesterday, Monday 28th of August 2023, OpenAI introduced their answer to business uses with all it’s privacy and security features: OpenAI Enterprise.
STOP…..
Are businesses looking for privacy and security as a service or merely basic features for demanded tools to sit on top? The answer is not so clear, because most businesses have NO CLUE on how to best capitalise on the tremendous power of a GPT neural network. Let’s analyze the business needs analysis to find out more. In a study conducted by the Harvard Business Review, it found only 3% of Companies’ Data Meets Basic Quality Standards.
The REAL use of any GPT model lies in it’s ability to read complex and vast amounts of data and then have this data ready for tasked queries, using natural language. Yet, it’s prowess depends solely on the type, structure, and relevance of data as well as the processing design applied for the intended output. So, a business would want to first and foremost have it’s own customer/consumer and operational data to be preprocessed and kept ready for capturing useful insights that would help a particular business develop enhanced customer relationships and optimise their processes. Where is the global large models like ChatGPT in all of this? Well, they are very powerful for the general user, covering a vast array of categories, but stops short of aiding any given business to understand and make use of it’s own in-house data.
So, the answer is not in offering privacy and security through better efficiencies. The business answer lies in creating much small Task-based GPT models that relate 100% to any given business, allowing it to capitalize tremendously on it’s own datasets within it’s confines.
Bigger is not always better, especially for Business AI.