In a rush to seize first-mover advantage with a nascent technological evolution, large corporations went all in on AI before AI was ready for prime time.
AI old schoolers like me kept shouting into the wind as early adopters began settling into one of three categories:
- Massive AI infrastructure investments: Many large companies made early, broad investments in the AI “new normal.”
- Overindexing on early LLMs: Coupled with the demands for immediate results, early adoption of large language models meant too much time, energy and dollars invested in chatbots and content creation. Today, we all see a lot of “AI user stuff” rather than more innovative “AI builder stuff.”
- AI as a replacement for labor: As the productivity promises of LLMs piled up, companies laid off huge swaths of employees, largely in roles that were increasingly viewed as obsolete, including the senior IT experts who could actually bring the AI builder stuff to life.
Many of those senior tech people are now available, and hiring one of them could pay huge dividends on the AI builder front, bringing predictive analytics, process automation and agentic development within reach.
DIVE DEEPER: Learn what you'll need to build a foundation for scalable AI.
It's Good That Small Businesses Adopted AI Late
As AI market share starts to settle and AI provider loss-leading evaporates, executing on those massive early investments is becoming more expensive; in some cases, even surpassing the cost of the human labor it replaced. AI coding costs will surpass the average developer’s salary by 2028, Gartner predicts, as token consumption surges and strains project budgets.
The good news is that most SMBs escaped this vortex. Being late to the AI party meant they didn’t make those mistakes with early AI adoption, such as mandating an entire organization use AI without guidelines, governance or training. In most cases where SMBs adopted AI, they tended to let their employees find their own productivity paths.
SMBs also didn’t have the bandwidth to replace key human-led business functions with LLMs. As early results increasingly show a lack of promised productivity with AI (especially for complex, critical business functions), there’s little for SMBs to undo. Maybe they need to unbolt a chatbot or roll back an unused AI-native service offering. SMBs were never great candidates to replace their human labor with AI to begin with, primarily because their employees usually serve more than one repeatable, replaceable business function.
Click the banner below to get small business insights delivered to your inbox weekly.
