I had the opportunity of attending IDC AI Roadshow 2024 on earlier this month on Thursday 9th May 2024 and saw a bunch of use cases for AI as well as partake in several interesting discussions.
It’s definitely an exciting time working in technology as advancements with AI are proving to be highly productive and advantageous to organisations that are experimenting with them. It is even going to the extent that organisations are restructuring several parts of their organisational structures to cater for AI, showing how much of an impact AI has been having especially here in the GCC.
I wanted to share some of my thoughts following on from the roadshow and give what were my main takeaways.
Generative AI & Bespoke LLM’s
The focus is still very much on Generative AI, and rightly so. This has been the main type of AI made available via tools such as ChatGPT, and it’s no surprise that organisations are training their own LLM’s based on their own organisational contexts. This then leads on to…
Use Case Wrappers
Once organisations have trained their own LLM’s, they are creating their own wrappers around it. This may be around document classification services or healthcare diagnosis services, however the core usage of the technology is similar. Only the use case changes.
What else?
Skillsets
Due to the usage of AI and it’s many use cases across organizations, it’s highlighted the importance of surrounding skillsets. This was mentioned as part of the keynote given by Ranjit Rajan, VP of Research (META) at IDC.
AI Governance
With the increased usage of AI, there are open questions around its governance and AI policies to ensure usage is done in compliant ways. Generally, AI compliancy and ethics is a very open and to be defined area, but organisations have to ensure that AI usage is inline with existing policies.
Data Handling
With the explosion of data inputs and outputs into and out of LLM’s, there is and will need to be a higher focus on data security and how to best use that data. Organisations are now dealing with more data than ever before, and need to make sure that data is of value rather than having data for the sake of data. Remember, data costs – those cloud storage fees are not cheap so there needs to be some thought put into what is the RoI for generating and retaining data.
The Future
It’s yet to be seen what other advancements will be made with AI. We are right at the beginning of the journey, and I feel that there will be several other key advancements in technology thanks to AI over the coming 5-10 years.