- VMware Product Release Tracker (vTracker)
- VMware Product Release Tracker (vTracker)
- VMware Product Release Tracker (vTracker)
- If your AI-generated code becomes faulty, who faces the most liability exposure?
- These discoutned earbuds deliver audio so high quality, you'll forget they're mid-range
Avoid generative AI malaise to innovate and build business value
Despite the promise generative AI holds for boosting corporate productivity, closing the gap between its potential and business value remains one of CIOs’ chief challenges. It isn’t for lack of effort, as recent research suggests.
Sixty-six percent of C-level executives are ambivalent or dissatisfied with the progress of their AI or GenAI efforts, according to Boston Consulting Group1. The research cited a lack of talent and skills to work with the technology (62%), unclear AI and GenAI investment priorities (47%), and the absence of a strategy for responsible AI (41%) as the top three obstacles.
Deloitte2 meanwhile found that 41% of business and technology leaders said a lack of talent, governance, and risks are barriers to broader GenAI adoption.
The GenAI FUD struggle is real
The concern even has some of the vendors behind leading GenAI platforms toning down their rhetoric. They fear that GenAI’s promise has outstripped its current value relative to its high cost, accuracy, and unclear productivity gains, putting them in danger of overpromising and underdelivering—a kiss of death in the tech sector.
Has GenAI finally fallen into Gartner’s dreaded “trough of disillusionment”? Perhaps. However such fear, uncertainty, and doubt (FUD) can make it harder for IT to secure the necessary budget and resources to build services. This could lead to more shadow AI, which could lead to more security threats and a wider attack surface.
As an IT leader, you’re in a unique position to help your organization avoid such malaise. You can innovate and protect your corporate data by running a private GenAI instance that affords you greater control over total cost of ownership, performance, security, and other critical factors.
But how do you get there? This playbook can help.
Reach consensus on strategy. You’ll build a cohesive strategy, centered around business use cases agreed upon by the C-suite, line of business leaders, and other key stakeholders. Include low-hanging fruit to bigger medium and long-term big bets for GenAI adoption. Plan the orchestration of people, processes, and technology within IT and be sure to incorporate governance policies and guardrails. You must also determine how you will educate employees on the dangers of shadow AI, as well as what GenAI services and best practices support safe, responsible use.
Assess your readiness. GenAI use cases require prudent infrastructure planning and deployment. Capturing the “as-is” state of your environment, you’ll develop topology diagrams and document information on your technical systems. Next craft a “to-be” blueprint of what you need to support your strategic vision, including targeted capabilities, future IT architecture, and talent required to facilitate the work.
Cleanse your data. GenAI requires high-quality data. Ensure that data is cleansed, consistent, and centrally stored, ideally in a data lake. Data preparation, including anonymizing, labeling, and normalizing data across sources, is key. You’ll also institute guardrails for data governance, data quality, data integrity, and data security. This will help you curb risks as well as reduce the likelihood of wasted efforts that can accompany dirty data.
Right-size your model(s). Creating private instances of pre-trained LLMs, such as Llama 2, can help you get up and running quickly while saving costs on inferencing. You may also choose to leverage retrieval augmented generation (RAG) to augment your model with domain-specific information, which may also reduce hallucinations. Low-cost proof-of-concepts can help you reduce the risk of overprovisioning. Ultimately, how you decide to right-size your models will depend on your use cases and the business outcomes you wish to derive from them.
Choose a workload location. Choose the operating environment that makes the most sense based on your business requirements. On-premises will allow you to customize your model and support it with hardware optimized to handle heavy compute and storage loads. Maintaining complete control over your own infrastructure and data will afford you more peace of mind as you embark on this transformational journey.
Pick the right partners. As you begin building, remember to start with the low-hanging fruit—something core to your business—that will help you gain maturity. High-risk (even if high-reward) use cases can hinder—or worse—your GenAI plans. And always, always keep a human in the loop to maintain organizational guardrails.
Remember: GenAI is nascent enough that there is no singular formula for success. And your IT organization may lack some of the resources, such as talent and tools, to facilitate your vision.
That’s where partners can help you choose your own adventure. Dell Technologies offers servers, storage, and client devices to run your models on-premises along with professional services to help you choose the GenAI path that’s right for your business.
Learn more about how Dell Generative AI Solutions.
