Getting the business results you want from generative AI isn’t easy — there’s a definite learning curve. What’s the fastest way to get up to speed?
Information Week explores this issue in Hire or Upskill? The Burning Question in an Age of Runaway AI. Leapfrog is glad to weigh in — we’ve been enthusiastically helping clients integrate innovative technologies into their IT ecosystems for decades, so here’s our take:
First, embrace the concept
Generative AI (GenAI) is more than a shiny new object. Leapfrog is seeing enormous AI interest (and concerns) across our client base because they want to leverage the possibilities while remaining secure.
As the Infoweek article states, “With the speed at which generative AI developments are advancing, it is no secret that virtually all brands will need to have some level of AI proficiency within their organizations,” says Ashu Dubey, Co-Founder & CEO of Gleen AI.
Talking with computers is different from talking with humans, so there’s a learning curve to developing effective “prompt skills” — it’s not like talking to Siri or Alexa. And it takes time to learn these skills.
Since the GenAI boom is so new, only some people have had the time to become prompt experts. These experts, called prompt engineers, are understandably in high demand and earn salaries as high as $300K or more. For many companies, the investment is worth it. Yet, democracy is inherent in GenAI — AI prompt proficiency is learnable. Soon, many people will be good at it, including, possibly, members of your own team.
That’s why Leapfrog’s short answer to whether to hire or upskill for GenAI is to do both. And to take advantage of AI integrations for platforms you’re already using, too.
Why hire a prompt engineer?
As the Information Week article points out, prompt engineers know how to break down the context of a question and express it in terms that AI models can understand.
But prompt engineering is currently a nebulous career field. Some prompt engineers have highly technical backgrounds and some have more general knowledge. When it’s time to hire a prompt engineer, look for someone who already understands your industry. That way you can generate the most meaningful exchanges between humans and machines and use the output to accelerate your business.
The more you can identify specific AI business cases where you’ll likely need prompt engineers, the more strategic your hiring can be.
Infoweek recommends that highly technical companies build prompting teams to work with prompt engineers. Some goals include:
- Combining the expertise of multiple domain-specific engineers
- Including input from individuals who understand different elements of your business
- Focusing the team on the five stages of prompt engineering: prototype, production, internationalization, polishing, and optimization.
While some experts believe prompt engineering will become a more significant hiring category over the next few years, others think it’s fleeting or that AI will soon learn to write its own prompts.
Why upskill your team?
In a recent survey, McKinsey found nearly 40% of respondents adopting AI expect to reskill more than 20% of their existing employees for GenAI. That’s a significant number of people on the AI-whisperer path.
If your primary goal is to incorporate AI tools into your tech stack, the Infoweek article states you might be better served by training current team members on platforms like ChatGPT instead of hiring prompt engineers. Leapfrog agrees.
To build GenAI proficiency in-house, take an organizational approach that embraces prompt writing as a fundamental business skill. Empower your people by:
- Providing training to help employees quickly advance their skills
- Encouraging those who show interest to develop their prompting skills even further
- Building prompt libraries that everyone at your company can use
If you need custom AI training solutions, consider having someone develop internal tools for you.
Why use enterprise integrations?
Since there’s not a pool of prompt engineers waiting for your job offer and it will take time to train your team, there’s another option you can leverage right now — enterprise integrations. Why not capitalize on the tools you already use?
The most popular enterprise integration for GenAI is Microsoft’s Copilot. It’s built into the operating system and works with your data in Microsoft 365 — Word, PowerPoint, Excel, Outlook, Teams, and other apps.
As with any integration, it’s essential to integrate Copilot correctly. Copilot automatically inherits the security, compliance, and privacy policies you’ve set in Microsoft 365, but it can also be deployed with specific rules about your proprietary information.
Leapfrog recommends that organizations begin to use Copilot and other AI enterprise integrations as long as they deploy and manage them in a security-centric way.
Why write clear AI policies?
At Leapfrog, we’ve configured our AI rules so that each team member’s work product can be shared with others at Leapfrog and are currently building out our AI boundaries. The Copilot and Microsoft Azure information protection features are in early release now — as a partner of Microsoft, Leapfrog is participating in the technology learning process so we can implement the solution for our clients as soon as possible.
Leapfrog highly recommends that companies develop an AI Acceptable Use Policy even in the early stages of GenAI. We’ve shared ours with our clients to guide them in advising their end-users.
Build your GenAI foundation
The faster your company can build GenAI into your systems, the faster you can monetize it. For most mid-market companies, this means both hiring experts and training your team. We believe that having a blend of specialists and generalists will enable you to get the most out of your GenAI capabilities.
And to get the most out of your AI, you need a modern, high-performing, and secure IT ecosystem — exactly what Leapfrog has been providing our clients for more than 25 years.
