GenAI: Boffins in clover

When Larry Ellison gave his keynote speech at Oracle Cloud World in Las Vegas last week, it would not have been shocking if he had expressed some scepticism about the hype generated by Generative AI over the past year. He’s seen it all, after all.

But not in the slightest. Ellison said GenAI does indeed mark the dawn of a new era. He spoke of how the “baby” ChatGPT is talking, with a multi-billion parameter large language model behind it. And how Oracle is using Generative AI to transform its own cloud technologies.

The previous week, in San Francisco, Salesforce’s Marc Benioff was also waxing lyrical about GenAI, at Dreamforce, invoking the toddlerhood of ChatGPT and Google’s Bard. In Salesforce’s case, they were touting a platform called Einstein 1 and a conversational assistant.

Oracle announced a slew of 50-plus GenAI features that will be rolled out in the next two quarters releases of its Fusion cloud-based applications suite. It also highlighted some GenAI aspects of the healthcare IT Cerner acquisition it made in 2021, including a voice-based clinical digital assistant.

It has been the same story at every IT conference in 2023. From Qlik World in Las Vegas in April, through ServiceNow’s Knowledge 2023 event in Vegas in May, and SAP’s Sapphire event in Barcelona in May, similar declarations of a new epoch have been made. Often by artificial intelligence boffins who have been slaving away on AI for years and years, in relative obscurity, and who are now, finally, in clover. Einstein himself could be said to be their avatar.

One socially interesting aspect has been the role of teenage children. “Man, you gotta check this out” has been the adolescent refrain at the breakfast table of many an IT industry C-level executive, and of many a CIO, or similar, at customer organisations.

As a quick reminder, Generative AI is a type of artificial intelligence technology that can produce content, including text, imagery, and audio. It uses large language models (LLMs), data models with billions or even trillions of parameters.

It is the text generation aspect that is critical, and profoundly novel. Masses of business report writing, including traditional Business Intelligence reporting, can now be pushed to the margins of knowledge work in companies and other organisations, allowing workers to focus on more creative tasks.

Computer Weekly’s sister analyst firm, TechTarget’s Enterprise Strategy Group has recently published some landmark research into enterprise use cases for Generative AI. It shows the  concept as ranking higher than sustainability and cloud migration as a strategic priority for enterprise IT, and discloses use cases like better chatbot dialogues, improved threat intelligence analysis, and faster code creation.

The message to CIOs is clear: put away any scepticism. This shit just got real.

Data Center
Data Management