Why 2025 Will Mark A Turning Point

By Eric Sydell

AI is the most disruptive technology in history, as it is supplanting human expertise and will — slowly and then quickly — diminish human control.

There are plenty of voices suggesting that artificial intelligence will hit a wall, or already has, due to its unending need for data, energy costs and regulatory issues. And to confirm this, AI marketing hype often paints a picture that is vastly inflated from its reality while developers constantly seek funding to continue their progress.

The fact is, AI’s growth shows no signs of slowing, and its impact will only continue to scale. For the past several years, there has been a divide between two AI camps: one led by physicists and technologists such as Max Tegmark and Elon Musk, and the other led by researchers including Joy Buolamwini and Timnit Gebru.

The former have argued that AI must be harnessed so it doesn’t take over the world, while the latter claim that the immediate threat lies in AI’s ability to amplify bias and invade privacy.

Both perspectives are valid and important. As Henry Kissinger, Eric Schmidt and Craig Mundie have pointed out, AI can either save humanity or destroy it. We must work together to build guardrails for AI to ensure it remains a positive force.

Yet, AI is already contributing to dehumanizing outcomes — including manipulative chatbots, deepfake pornography and biased emotional assessments.

Proceed with caution

It remains uncertain whether global leaders can summon the will to harness a technology that is difficult to understand, massively transformative and deeply entangled with political funding. As GenAI gained prominence, the tech hype machine went into overdrive with the amazing capabilities of this technology.

Now, however, we have likely entered the Gartner Trough of Disillusionment with GenAI, as the hype has started to wear off and organizations are finding it difficult to use at scale.

We are getting there, though, and in 2025 we can expect to see continued progress across several fronts.

Specialization over intelligence: Foundational models will evolve with new features, but their root intelligence may reach a limit. Expect to see more useful add-on features to extend their capabilities.

Finding value in specific use cases: Experts are developing tailored AI solutions, giving large enterprises a competitive edge by leveraging GenAI for specific use cases.

• Next-generation SaaS applications: New tech providers are building scaled tools on top of GenAI (referred to as cognitive architectures), creating opportunities for small and mid-sized firms to catch up with larger players.
• Balancing AI autonomy with oversight: The AI hype cycle is shifting to agentic AI, which combines GenAI with traditional coding for autonomous tasks. Careful oversight is needed to control these agents’ actions and access.
• The growing digital divide: A digital divide is emerging between those using AI for knowledge work and those who aren’t, giving the former a significant productivity edge.
• Organizational realignment and job shifts: AI will automate more intellectual labor, driving job losses, role shifts and the creation of new positions in management in 2025.
• Shift in regulations: New regulations may target tech firms that censor users’ posts, with misinformation, but broad AI regulation is unlikely under the new administration, which will focus more on innovation and AI dominance.

Going forward, organizations must carefully evaluate the potential impacts of AI on their business — both the opportunities and the risks — and adapt their structures to leverage its benefits while mitigating downsides. These risks could dehumanize employees, leading to dissatisfaction and lower performance.

Organizations should be cautious of AI hype and remember that compliance alone won’t guarantee meaningful results. Ensuring responsible, effective deployment will likely lead to compliance by default.

Eric Sydell, PhD, is co-founder and CEO of Vero AI, a firm dedicated to understanding and harnessing the power of AI and all algorithms for the benefit of human society. Sydell is an industrial-organizational psychologist, entrepreneur and consultant with more than two decades of experience working in the recruiting technology and staffing industries. An expert in AI, machine and deep learning, psychometrics, and their practical application in hiring, he regularly writes and speaks on these topics, both in the media and at academic and industry conferences, worldwide. Sydell is the co-author of the book “Decoding Talent: How AI and Big Data Can Solve Your Company’s People Puzzle,” from the Fast Company Press.

Illustration: Dom Guzman