To prepare effectively for the future, Mehdi Paryavi argues that businesses should stop viewing AI as a single phenomenon and start focusing on agentic AI as distinct from today’s generative tools
Reports about the AI boom are now your quotidian fare. You can easily find analysts proclaiming, “X percent of companies are now using AI,” “X percent of enterprises have prioritised AI,” and the like. You can also see emerging talk of an AI bubble, with fears of a repeat of the dot-com crash of 2001.
So, what should you do? Does your company have a dearth of AI talent and “AI readiness”? Last year, the organisation I lead, International Data Center Authority (IDCA), completed a comprehensive report – The Global AI Report 2025 – on AI readiness among nations. We measured each nation’s underlying IT infrastructure, its progress with respect to sustainability, its governmental effectiveness and stability, and the degree to which it had a roadmap for AI adoption. You can do the same on an organisational level.
We can assume your organisation has a strong, resilient IT infrastructure and robust customer-facing services, whether along your value chain or your direct, retail customer experience. We can also assume your company has had sustainability in mind for some time, and that your executive team and directors operate at a more-or-less effective level. But we cannot assume that your executives, managers, and overall workforce know precisely what AI is today, what it can do, and whether or not it can present a genuine threat to existing jobs. Are you really ready for AI?
Will it take my job?
“The computer is going to take your job” has been a staple of scary headlines and reports for decades. Some of us will remember when the personal computer was feared and fought by IT departments who thought this new invention would take control from them, kill their jobs, and cause anarchy. Similar fears were expressed when the World Wide Web delivered the internet to anyone with a connection.
Now these fears have returned with AI. Within this context, IDCA research has found a worldwide need for 101 million new IT-related jobs by 2030. Each nation has its own share of this need. Those who meet it will progress, and those who don’t will fall further behind. At the same time, AI threatens to automate and eliminate any job that requires minimal skill or represents minimal value to the organisation. The answer is nothing new or novel. The answer lies in the proven commitment to training and certification, and the upskilling it provides. A worker equipped with an effective AI service outperforms a worker without one and outperforms a piece of AI software trying to work by itself.
Having said this, many jobs will be erased and completely wiped out. In fact, most computational, analytical, and even decision-making jobs will be replaced. The question is, who is prepared to embrace the future with AI, and who is still in denial, hoping that this bubble will soon burst. If you stay where you are and as you are, yes, AI will take your job. But if you adapt and adopt, chances are you will develop an extended reach and audience for your specific set of skillsets, using AI.
“GenAI does not paint the big picture. Agentic AI does”
Agentic AI is the key
Furthermore, AI does not describe a single product or service, and it continues to evolve. Today’s focus on generative AI (GenAI) services such as ChatGPT, Anthropic’s Claude, and others will soon be augmented, or superseded, by emerging agentic AI services. Where GenAI uses large software models to predict – often incorrectly – how it will answer your questions, agentic AI is goal-oriented and can improve itself over repeated experiences.
By focusing on agentic AI, we are also suggesting that current GenAI platforms and services are not the endpoint of AI. Perhaps the big developers – we’re especially looking at you, OpenAI – will solve the propensity of GenAI to hallucinate, or maybe at some point concede that the current development approach around large-language models (LLMs) is a dead end. The latter development will indeed rock AI’s world, and maybe rock the stock market for a while.
But GenAI does not paint the big picture. Agentic AI does. To get a handle on agentic AI, we suggest employing a combination of the old and the new. Do a thorough analysis of all your business processes and how your organisation is structured. If you already know what the main pain points are, look again to prioritise the top several things you want agentic AI to improve. Survey your workforce’s fear level about AI, and please, don’t threaten everyone with losing their jobs to a piece of software.
We trust that enlightened readers know better and will work with everyone in their organisation to upskill jobs, improve those recalcitrant processes that impede your business, and benefit from a robust application of agentic AI.
Beyond agentic AI lies the epochal quest for an artificial general intelligence (AGI), something that we believe is not relevant to your organisation at the moment. You can certainly be aware of what all the researchers are doing and saying. Grappling with AGI and its ramifications will be tomorrow’s problem for your business.
But the heart of AI today for your organisation is agentic AI. Focus on it, debate it, and you may end up rejecting its widespread use. But an informed decision to do so is much different than simply ignoring it, either out of fear or arrogance, or even worse, pre-emptively announcing mass layoffs because of AI in all its forms.
About the author
Mehdi Paryavi is the Chairman and CEO of the International Data Center Authority (IDCA), the world’s leading Digital Economy think tank and prime consortium of policymakers, investors, and developers in AI, data centres, and cloud.