Industry executives and experts share their predictions for 2024.  Read them in this 16th annual VMblog.com series exclusive.

Two Under-the-Hood Predictions Around GenAI in 2024

By Lalit Ahuja, Chief Product and Customer Officer, GridGain

While 2023 will be remembered as the year of generative AI (GenAI), we have barely begun to experience its impact on businesses and society. For GenAI to truly transform how we work and live, it must first be securely operationalized, and to do that, enterprises must understand that GenAI produces more relevant outputs and works better with more context. That is, it depends entirely on massive amounts of data, often including proprietary data that is specific to the user. And every piece of information we give to the GenAI models allows it to learn more about us. As such, the success or failure of every business's GenAI initiatives will depend on the willingness of the business to share its relevant data and find a way to do it in a controlled and secure fashion.

As a result, predictions around the evolution of GenAI will invariably be tied to the evolution of governance frameworks around data management, the exact use of the GenAI models, and the associated security. With that in mind, I offer two predictions related to key enterprise developments in 2024.

Securing GenAI

Every CXO today must be thinking about how, and how much, to leverage GenAI in operating their business and how the technology can create differentiated value for customers. However, if that "X" is working in a legal or InfoSec role, then they will spend 2024 scrambling to find the right balance between allowing the growth of GenAI and limiting its risks by instituting critical frameworks and safeguards.

This means that the massive adoption of freely available LLMs and the use of ChatGPT will quickly give way to enterprises enforcing their own constraints on usage. First, a business must be able to ensure only the right people in the right roles have access to the information going in or out of the models. Further, as models consume all the information fed to them in the form of questions or content for further refinement or analysis, they are learning more and more about the user, consumer or business providing the information. 

Enterprises must understand the strategic, legal and security impact of who has access to that model. As a result, they will start preventing unfettered use of the technology - or prohibiting it altogether - to protect corporate IP, customers and employees. In early 2024, I expect to see a rapid rise in actions such as blocking domain names like openai.com and implementing other such controls. 

Modern Data Architectures

As noted above, data scale, speed and security are foundational to the success of GenAI initiatives. Ensuring these capabilities will require a modern data architecture that recognizes the multi-dimensional nature of today's data processing needs. This will lead to two key developments in 2024.

First, data integration hubs (DIHs) and DIH technology providers will try to expand to cover a bigger chunk of the data value chain by providing additional capabilities. Today's DIHs are a great foundation for a modern data architecture, but they lack key capabilities, such as analytical/AI/ML processing and persistence. They must also solve the challenge of latency as data is moved in and out of the hub. To do this, DIHs will necessarily gravitate towards becoming "unified" data processing platforms, as defined by Gartner.

Second, to build a modern data architecture, there will need to be a level of consolidation at both the logical and physical architecture levels. This means it won't just be a consolidation of technology components. There must also be a consolidation at the underlying application and service provider levels. This will also lead to increased M&A and consolidation activities at the data solutions and solution provider levels. Expect this activity to ramp up throughout 2024.

Conclusion

We are clearly at a societal inflection point around the application of GenAI, which makes it fun and exciting to make bold predictions about how GenAI will change everything. But for any of that to happen, enterprises must first do the hard work of operationalizing it - without sacrificing security, privacy or compliance. It won't always be easy, and it probably won't happen fast, but it's the necessary groundwork for success.

##

ABOUT THE AUTHOR

Lalit Ahuja is the Chief Product and Customer Officer for GridGain, a provider of a leading unified real-time data platform, where he is responsible for ensuring customers gain the most value from their GridGain investments.