Let’s get one thing straight: The world is in the throes of an AI frenzy, with generative AI and ChatGPT making headlines almost daily. Everywhere you look, you’ll find entrepreneurs looking to start the next big AI company, executives keen to leverage AI for their businesses, and investors eager to bet their money on AI ventures. I’m all in for the power of large language models (LLMs), and I’ve experienced first-hand their potential in boosting productivity.
Business and AI: Beyond the Hype
In today’s landscape, business and technology are inseparable. When a new tech trend hits, businesses expect it to streamline operations and push performance to the next level. AI is no exception. But, it’s important to remember that AI isn’t a silver bullet that magically solves all business woes. A well-oiled business machine will continue to thrive with or without AI, just as a poorly managed one won’t suddenly prosper just because it embraced gen AI. Successful AI implementation isn’t just about the tech’s performance; it also requires savvy management, akin to handling any other aspect of business operations.
The Bubble and the Burst: The AI Hype Cycle
As with any new tech, gen AI is on a roller coaster ride known as the Gartner Hype Cycle. Applications like ChatGPT have thrust gen AI into the limelight, leading us towards the peak of inflated expectations. But, as the saying goes, what goes up must come down. We’re looking at a future trough of disillusionment as the initial excitement fades, experiments flop, and investments go belly up. This downturn could be triggered by several factors such as tech immaturity or ill-suited applications. Yet, I’m spotlighting two common disillusionments that entrepreneurs, executives, and investors need to understand to navigate the choppy AI waters successfully.
Does Generative AI Really Level the Playing Field?
One popular belief is that gen AI democratizes business. It’s easy to buy into this narrative when you see millions using gen AI tools for various tasks, making English the new coding language. While there’s some truth to this for specific applications, it’s not the whole picture. Gen AI is still grappling with tasks requiring deep domain knowledge. Remember when ChatGPT generated a medical article littered with glaring inaccuracies and botched a CFA exam? That’s what I’m talking about.
Domain experts may possess vast knowledge, but they’re not necessarily tech-savvy or well-versed in the workings of gen AI. For instance, they might struggle to prompt ChatGPT effectively to get the desired results. Moreover, the rapid developments in AI make foundational LLMs a commodity. To stand out, an LLM-enabled business solution needs to offer something unique, like access to valuable proprietary data or specialized expertise.
Existing businesses often have this domain-specific knowledge, but their legacy processes could impede swift gen AI adoption. New businesses, on the other hand, have the advantage of starting fresh and fully leveraging the tech’s power. However, they must hit the ground running to amass vital domain knowledge. The challenge then is to empower these domain experts to train and supervise AI without needing to become AI experts themselves, while capitalizing on their domain data or expertise.
Key Considerations for Generative AI Adoption
Let’s be clear: gen AI has its limitations. We need to leverage its strengths and sidestep its weaknesses. Here are some crucial technical considerations for those mulling over investing in gen AI. If you’re thinking about building in-house solutions
