Naveen is the CIO of Zutshi databricks,Thank you for reading this post, don't forget to subscribe!
In the year since Generative Artificial Intelligence (AI) found its way into the broader public conversation, its application in enterprise technology has only grown – and so has the excitement among leaders.
The current wave of innovation has just begun, and the work the technology industry is doing now will reveal how AI will impact our lives and work in the long term. Leaders are currently addressing their short-term priorities with generic AI, which will bring a number of trends and challenges in 2024, including:
Leaders must weigh cost and operational considerations.
Before anything significant happens with generative AI in their companies, leaders must decide: buy or build? Big or small model? Open source or proprietary?
Will be associated with some big models. Others will likely choose to invest in smaller models based on internal data. Some people will find that they need large language models (LLMs) trained on billions of parameters, but still want to enhance those models with their data to ensure they are getting the most from their investment. . Some companies will opt for both, and some will take a wait-and-see approach, with others watching from the sidelines to see what is most successful.
On the seller side, an important, still debated issue revolves around pricing. There is no standard or baseline for pricing of AI features, and some vendors are considering extremely high pricing per user per month. As some vendors pass on rising GPU and research and development costs to their customers, CIOs are still being tasked with cutting expenses and will be forced to charge ancillary fees for the core product as well. , which they are accustomed to.
To address this concern, more CIOs may choose to create their own models. The market is so new that purpose-built models for specific tasks or industries may not be available yet, and security and data protection are major concerns driving companies to create or improve their own models, at least in the short term. .
Companies will face hurdles related to data quality, model reliability, access, and governance.
Many companies are experimenting with LLM, but few of those solutions have gone into production. Leaders want to get data out to the public but haven’t yet figured out how. Many software companies—including my company, Databricks, and Microsoft, Glynn, Replit, and GitHub—are enhancing their products by implementing generative AI capabilities. We are also seeing a growing number of other vertical industries implementing generic AI offerings tailored to their industry use cases.
The main hurdle to “doing generative AI” revolves around implementing the technology in a safe, effective, and reliable manner. Leaders are grappling with data issues, particularly access to and governance of an organization’s trove of data.
For example, in a Workday survey of more than 2,300 executives globally, only 4% of respondents described their data as completely accessible – reflecting a huge gap between many companies. want What to do with generic AI and what they are actually equipped to do.
The accuracy and reliability of what is generated by LLM is also an important research and industry challenge that must be addressed. Departments that require a higher level of accuracy – for example, legal, finance and compliance – will adopt generative AI tools more slowly than other functions, but will still see benefits. However, to deliver true value to these teams the models must be more reliable – consider the degree of accuracy required for a non-technical lawyer or financial analyst to do their job well with the help of generic AI.
While accelerated engineering, retrieval automated generation (RAG) models and fine-tuning can reduce hallucinations, they do not eliminate them – so I expect there will be more innovations to address this main concern.
In the coming year, data leaders will be focused on overcoming these fundamental challenges. We will likely see more companies consolidating their data tools and doubling down on efforts to improve data quality and control. A recent MIT Tech Review study conducted in partnership with Databricks showed that nearly three-quarters of senior technology executives have adopted the lakehouse architecture, and many more are planning to do so.
where do we go from here
The work – some experimental, and some successful – the teams are doing now will reveal where companies will get the most immediate value from generative AI and, subsequently, where they will invest the most money and time. But none of this will be possible in the near future without a new focus on data integrity, quality and governance.
Millions of knowledge workers—both technical and business users—will use various forms of AI-powered assistants in 2024 and beyond. The sheer number of people using these devices gives the technology industry a large sample size to gain insight into how and to what extent AI can successfully assist in various tasks. This will ultimately help inform future perspectives, success factors, and use cases for generative AI.
Like all great technologies, generative AI will evolve (and even falter) before reaching its full potential. We still have a long way to go with generative AI, from increasing its efficiency and accuracy to reconciling the necessary regulations with innovation. One thing is certain: the coming year will be integral in defining what an AI-powered future looks like and how it will ultimately change the way organizations operate.
The Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs, and technology executives. Am I eligible?