Roundup: Why AI Development Will Slow – a Bit – During 2025
The use of AI will continue to grow but technology limits, the business environment and stuff thrown against the wall will act as a brake.
Welcome to WorkforceAI’s new weekly Roundup. Each week, we’ll share the industry’s vital news along with analysis of what’s happening in AI, HR and the AI business. Paid subscribers will continue to receive breaking news updates as they occur. Our daily Bulletin has retired. Do you have areas you want us to cover? Let us know.
After two years of sometimes manic development and discussion, the world of AI is ready to take a breath.
It’s not that we’ve saturated interest in AI. It’s certainly not because people aren’t finding AI useful. On the contrary, the number of businesses pursuing some kind of AI strategy continues to grow. According to McKinsey, more than two-thirds of survey respondents in most every region reported their organizations use AI. Meanwhile, the number of organizations adopting AI rose from 50% to 72% in 2024.
People are certainly finding ways to put AI to real use in areas like data analysis, automation, personalization and support. HR departments are applying AI to talent acquisition, learning, performance management and workforce planning. They’re also looking forward to new areas where AI could prove useful, like skills management and coaching.
During 2025, however, developers will face several challenges to their pace of change.
First, limits to today’s technology will become more apparent. A growing number of experts believe developers won’t be able to sustain today’s forward momentum. One of the most mentioned notions is that improvements to large language models are "plateauing." For instance, the advances of new versions may not be as dramatic as what we’ve seen in the past. Case in point: OpenAI’s Orion hasn’t displayed the same kind of improvements seen between earlier versions, like GPT-3 and GPT 4.
Then there are economic considerations. Many industry observers believe development costs will rise along with many of the business’s foundation blocks, such as the cost of AI chips, training and data centers. At some point, they believe, those costs, and the cost of product development, will rise beyond the point where investors will see a reasonable return.
Another thing to consider: The cost in playtime. Applying AI to new media, like video, holds great promise, but currently offers limited value. Many AI engines cap a video’s length at around 20 seconds, which limits their applications. Quick videos may be useful in some ways, such as emphasizing a point in onboarding materials. However, tools like OpenAI’s Sora and Google’s Veo 2 feel like applications that have been tossed into the market as a tease, designed to develop a user base for full products once they’re rolled out.
Finally, what would a look-ahead be without a mention of regulation? Already, some applications have been held back from the EU because of privacy concerns, and jurisdictions around the world are comparing their regimens to others. Lawyers and legislators will be busy, but those on the product and development side may find themselves held up at regulatory forks in the road.
Advertisement: Looking for the perfect HR software? SelectSoftware Reviews offers tailored guidance to help you choose the best HRIS, ATS, and payroll tools for your organization! No costs, no obligations, just great advice from people who understand HR technology. Visit SelectSoftware Reviews.
The News
Podcast: Kyle Lagunas of Aptitude Research on AI, Shiny Objects and HR
AI’s everywhere, but what changes is it bringing to HR’s fundamentals? How is it – and where is it – making itself felt, and how can CHROs sort out the reality of what’s going on when they talk to vendors, who want to hype? My guest is Kyle Lagunas, head of strategy and principal analyst at Aptitude Research. We’ll get into all that, and more, on this edition of PeopleTech. [Listen here]
Will AI Help or Hurt Workers? One 26-Year-Old Found an Unexpected Answer.
New research is a step toward figuring out what AI might do to the workforce, by examining AI’s effect in the real world. One finding: w research shows AI made some workers more productive –but less happy. [WSJ]
Payroll Service Market Anticipates Strong Growth Amid Rising Automation Demand
The cloud-based payroll software market will see a significant increase in market value as employers rely more on automated systems to manage payroll processes. The industry is evolving rapidly, driven by advancements in technology and shifting workforce dynamics. [PRN]
Virtual Vocations Survey Uncovers the Impact of Silence in the Hiring Process
Candidates navigating the remote job market are being left in the dark by potential employers. More than half, 57%, said ghosting takes place frequently or almost always during remote job searches. About two-thirds, 67%, were ghosted after submitting applications, indicating that communication breakdowns are most common in the early stages of the hiring process. Some 38% either withdrew or considered withdrawing from the hiring process because of poor communications. [PRWeb]
iCIMS Appoints Nikki Grigsby as New Chief Customer Officer
iCIMS hired Nikki Grigsby to be chief customer officer, responsible for customer success, professional services and technical support. Grigsby has experience in high-growth technology companies and has improved the customer experience and client success efforts for global enterprise organizations. She succeeds Diane Fanelli, who is retiring. [iCIMS]
Poll
Image: iStock