- AI investments are sky high right now, but the workforce isn't yet ready for what's to come
- Enterprises need workers with the skills necessary to deploy and run AI but must also train their workforce on how to use it
- AI is also poised to change how jobs are performed, analysts told us
$315 billion. That’s how much Alphabet, Amazon, Microsoft and Meta are planning to spend in 2025 building data centers to fuel the artificial intelligence (AI) boom. And it makes sense, given generative AI is helping drive monster revenues for all four companies. But there’s another piece of the puzzle that has to fall into place for the expected returns on their investment to materialize: the workforce.
To help ramp AI adoption, enterprises need a workforce that is both willing and able to embrace the technology. And right now, at least, that’s a problem.
According to a recent Ernst & Young survey of 500 enterprise leaders in the U.S., 90% said employees are encouraged to use AI for day-to-day tasks. But more than half (53%) of respondents noted employees are feeling overwhelmed or exhausted by rapid AI developments, and 65% said it’s hard to keep workers motivated to use AI.
In a separate study conducted by the firm, 80% of workers said they’d feel more comfortable with AI if they had access to more training opportunities.
“Like any new technology, it will most definitely take time for users to figure out the best way to use AI to help them be more productive,” J. Gold Associates Founder Jack Gold told Fierce.
Gold said that for all he sees companies investing big in AI, that money isn’t necessarily being spent on making it easy to use. True, there will always be a learning curve for new technology, he added, but few companies appear to be dealing with how to train employees to actually effectively use the AI tools that are appearing on their desktops.
“Just giving someone an AI agent does not automatically mean they can use it to best effect,” Gold continued. “AI, like Google Search, will only give you good results if you know what/how to ask it. AI will get better with that and lessen the user burden over time, but it will have to undergo a learning curve itself to do so.”
There are plenty of consultancies and edtech platforms like Simplilearn out there peddling AI crash courses and certifications. As Simplilearn noted in a blog, there’s a need for upskilling to even implement AI in the first place. That is, enterprises need engineers to organize their data, build the compute and cloud environments to run AI, test applications and manage ongoing operations.
Mark Moran, Simplilearn's CMO, told Fierce "Most of the IT workforce does not fully have the AI expertise to keep pace with hyperscalers’ rapid expansion. While there is growing awareness and investment in AI training, the skill gap remains a significant challenge, especially in AI model development, MLOps, cloud infrastructure, and AI governance."
But then there are the vehicles through which models are consumed. As Gold noted, a model can be great, but if the user interface is trash, it can “create a real problem for users.”
AI job-pocalypse?
But beyond requiring upskilling, it’s worth asking in what other ways AI will change jobs. And answering that is a bit more complicated than you might think.
Moor Insights and Strategy VP and Principal Analyst Jason Andersen told Fierce questions about skillsets and adoption are one piece of a much larger puzzle when it comes to thinking about AI forecasts. The other pieces, of course, are predictions relating to the development and evolution of training and inferencing. Basically, how AI is used will influence training and adoption and vice versa.
Andersen said he expects more businesses will adopt AI for security and data reasons, which in turn will require training “much more complex agents that could marginalize or even eliminate some job functions.”
This is already starting to play out in the customer service and marketing segments right now since initial use cases are, for instance, focused on using AI to deflect more care calls and more readily generate marketing content.
“So, the question in my mind is how autonomous will agents become in the next 12, 24, and 36 months? If they are fairly autonomous, then what roles will knowledge workers occupy and how will they interact with the AI?” Andersen mused. “Right now, we invoke the AI, but in the future in some processes the AI will invoke us as an escalation or approval mechanism.”
But don't worry, most don't expect AI to replace people - at least not entirely. Why? Well, as Simplilearn's Moran put it, "While AI can automate tasks and enhance efficiency, it lacks creativity, critical thinking, and ethical judgment. We’ll need skilled professionals to train, manage, and innovate with AI. The future is about collaboration rather than attaining complete self-sufficiency."
Update 5 pm ET: This story has been updated to include comments from Simplilearn.