Generative AI is the biggest breakthrough technology in years, democratizing information creation for the masses. As genAI caught fire in 2023, many organizations rushed to test and learn from the technology and harness it to grow productivity and improve processes.
Yet ironically, genAI’s potential has created a paradox within organizations.
On one hand, 76% of IT leaders agree that genAI will be significant if not transformative in boosting productivity, streamlining processes and reducing costs for their organizations, according to a Dell survey.1
On the other hand, most organizations are moving too slowly to advance genAI initiatives, suggests recent research.
Ninety percent of C-suite executives are either waiting for genAI to move past its hype cycle or experimenting with it in small pilots because they don’t believe their teams can navigate the transformational change posed by genAI, according to Boston Consulting Group.2
Key challenges include a shortage of talent and skills (62%), unclear investment priorities (47%), and the lack of a strategy for responsible AI (42%), BCG found.
Dismal Data: High Promise, Low Adoption
BCG isn’t alone in clocking these concerns.
Only 27% of business leaders surveyed by Accenture3 say they are ready to scale up genAI initiatives, with 44% saying that it could take more than six months to do so and take advantage of potential benefits.
In fact, few organizations are deriving value from genAI today, with only 5% implementing the technology in production and at scale and only 50% having the requisite talent to implement genAI well, according to Wavestone research.4
Additionally, while 63% have guardrails in place to use AI safely, these organizations worry about its role in misinformation, ethical bias and job loss among other risks, Wavestone found.
Such bleak statistics suggest that indecision around how to proceed with genAI is paralyzing organizations and preventing them from developing strategies that will unlock value. Even as organizations plan to boost spending on genAI in 2024.
Fortunately, a playbook exists to help organizations manage the enormity of the technological change genAI presents, while cutting through the fear, uncertainty and doubt to jumpstart business growth. The playbook marries organizational readiness with governance and iterative development.
Craft a strategy, build consensus
GenAI is a transformational journey. As IT leaders, you must work with your business peers to craft a strategy that orchestrates people, processes and technology.
The right strategy includes quick wins in the short term and bold bets for the long term. You’ll install and train staff to harness the talent and tools to execute the vision.
What does that vision look like? You’ll align desired near-term and future states to test-and-learn pilots as well as potential production projects.
You’ll also convene workshops articulating strategy and build consensus around what organizational readiness will look like. Key tip: Strategy, not technology, is critical for success with genAI.
Pristine Data is Key
Some organizations may let their data get stale. You’ll clean it up and make it sparkle. Prepare the data through anonymizing, labeling and normalizing across data sources and create guardrails for governance, quality, integrity and security. High-quality data will be the oil that makes your models hum.
Build vs buy vs open source
Building genAI models is hard for the most well-heeled organizations so you should consider open-source LLMs and tools as alternatives to building or buying expensive commercial solutions. Some of this will hinge on what you want your model to do. Right-sizing models is also important, as larger models require more servers, storage and energy.
Consider a small language model running on-premises as an alternative to public cloud LLMs. This will reduce infrastructure overhead while helping you manage costs.
Execute the strategy
You’re ready to begin building. To better tailor your model using enterprise-specific data, leverage retrieval augmented generation (RAG). Baby steps to the production projects, which could include a new enterprise search solution for employees, or perhaps a product search catalog that helps speed decision-making for customers.
Data analytics collected every step of the way will help assess performance and find blind spots that can hinder progress. Adjust and pivot on the fly as necessary.
The post-mortem after party
Project post-mortems are a vital part of the organizational learning process. Take the analytics you’ve collected and assess what went right and what went wrong.
How can you improve? And how can you position your organization with its best foot forward as you prepare new genAI initiatives? Ask and answer these questions honestly. It will be good practice for when you explain your work to your CEO and the board.
One more thing
Even a flawless orchestration of this playbook won’t guarantee success. Hedge your bets by choosing a partner you can trust.
The right partner will provide high-performing infrastructure, adoption frameworks and reference designs, as well as integrations with key ecosystem partners to help you build.
Dell Consulting Services offers the Generative AI Accelerator Workshop, a half-day interactive session with business and IT leaders who will help assess your organization’s genAI readiness.
1 Generative AI Pulse Survey, Dell Technologies, Sept. 2023.
2 From Potential to Profit With GenAI, Boston Consulting Group, Jan. 2024.
3 Accenture 2024: Pulse of Change Index, Accenture, Jan. 2024.
4 2024 Data and AI Leadership Executive Survey, Wavestone, Jan. 2024.
Read More from This Article: Generative AI readiness is shockingly low – these 5 tips will boost it