As I write from our offices in Sheffield, across the Atlantic, IBM is sharing their latest developments at their annual customer and business partner conference, IBM Think 2024.
As with all these occasions, the announcements at IBM Think come thick and fast, and if you want to catch up with them, the sessions can be found on the IBM Think 2024 website.
This year they are focusing on Generative AI with the Watsonx platform, Automation, hybrid cloud, scalability, and governance. There are a plethora of use cases and a variety of case studies on show—plenty to pique the interest of technophiles, financial whizzkids and creatives alike.
Having just watched the opening keynote (now available here), two things stuck out for me.
Enhancing customer experience
Firstly, Arvind Krishna (Chair and CEO of IBM), identified three areas where AI will have a rapid impact: customer experience, coding and digital labour. And it was the first of these that got my ears pricking up.
In an example given of how a telecoms company had incorporated AI into their customer experience, they reported that they were now able to handle 800,000 customer calls per month, that AI assistance reduced call waiting time by 30%, and that they were able to boost customer satisfaction by 40%. I don’t know about you, but that sounds like a good argument for implementing an AI customer service assistant using IBM Watsonx Assistant.
Digital labour and expert systems
The second thing I could relate to was the concept of digital labour. They offered a brief description of InstructLab, a method for generating high-quality synthetic data for training LLMs to amplify the unique knowledge and data that an organisation uses to train its bespoke data models.
It is a taxonomy-based tree of knowledge derived from an ordered form of data ingested from resources provided by the organisation. This data is used to train the model. However, this data set may not be large enough to populate the model fully. Therefore, synthetic data is generated to fill in the gaps. As more original data is generated by the organisation, the model can be further trained to provide richer outputs.
DeeperThanBlue’s Expert Systems
This caught my attention because of the Expert System use case that DeeperThanBlue has been developing to assist vehicle technicians with maintenance and repair procedures. We posted a demo of this system, which can be seen below.
If you do get the chance to look at the IBM Think conference content, you might observe that many of the use cases feel a bit out of reach for most businesses. It is only natural that IBM wants to demonstrate their solutions at their greatest potential. It is, however, important to recognise that the applications are scalable and accessible to all. Businesses may not be ready to commit themselves hook, line and sinker to Generative AI, so we’re here to ease the transition in a manageable way.
Generative AI is here to stay and will have an inevitable impact on business worldwide. It is those that act fast and scale up quickly that will benefit the most.