When it comes to genAI models, size plays a crucial role (and smaller might be more beneficial)

Brown portrayed data scientists as rare creatures, similar to unicorns – unique and often requiring compensation like that of a legendary being, he added, “as they well deserve.”

Many businesses can only afford to hire a limited number of data scientists, either due to the scarcity of qualified individuals or the expenses involved in their recruitment, leading to barriers in effectively training and fine-tuning the model, according to him.

A shift towards a hybrid approach?

As per Brown, CIOs have been gradually moving away from monolithic technologies, starting from transitioning from UNIX to Linux in the early 2000s. He suggests that AI is currently at a comparable turning point and advocates for a hybrid strategy, akin to that of a hybrid cloud, as the most favorable for implementing AI models. While the bulky and somewhat undefined LLMs are receiving attention presently, he envisions that the upcoming IT environment will consist of 50% applications and 50% SLMs.

“Data is scattered across various locations, whether it resides on-premises, in the cloud, or at the edge. Consequently, data is inherently hybrid, and since AI must operate where your data is stored, it also needs to be hybrid,” Brown remarked. “In essence, we frequently advise our clients and partners: AI is the ultimate hybrid workload.

“In essence, a CIO will have as many AI models as there are applications. This implies that training must be accelerated, tuning must be expedited, and expenses must be minimized. The solution to this challenge can be found in open source,” he elaborated. “Just as it enabled the democratization of computing, open source will do the same for AI; in fact, it is already heading in that direction.”

About Author

AndyC

Andy Curtis is an award-winning security consultant, researcher and public speaker. He has been working in the computer security industry since the early 1990s, having been employed by state and federal government, leading healthcare and banking providers across three continents. He has given talks about computer security for some of the world’s largest companies, worked with law enforcement agencies on investigations into hacking groups, and is a regular voice on TV and radio explaining IT security threats.

See author's posts