There has been speculation for a while about whether investment in artificial intelligence constitutes a bubble primed to burst. Recently, however, experts on AI development have been issuing stronger warnings.
Seeing Diminishing Returns
Ali Chaudhry is an industry voice to take seriously; he is a research fellow at University College London (UCL) and serves on Oxylabs’ AI/ML Advisory Board, which is concerned with AI and machine learning applications to web intelligence. Chaudhry, among others, predicts that in 2025, investors will be forced to lower their expectations for what general AI and large language models (LLMs) are capable of doing. The issue, according to Chaudhry, is that the technology is not yet as scalable as its developers have often suggested. “I think we will see diminishing returns in the capabilities of LLMs,” Chaudhry cautions. “Some AI labs are already hinting that scaling laws are not as effective anymore.”
Safety and Ethics Concerns
Chaudhry also predicts that increased government regulation and rising public concern that AI could become dangerous will dampen enthusiasm for general AI. For example, in a series of controlled tests in December, a new LLM was caught attempting to deceive developers to avoid being shut down. “2025 will be very important for AI safety,” Chaudhry says, “and we will see a lot of work (technical and non-technical) in this space.”
Adi Andrei, the cofounder of Technosophics and another member of Oxylabs’ AI/ML Advisory Board, has 25 years of experience in AI, machine learning, and data science, including at NASA, and he is similarly skeptical about the immediate future of general AI. Andrei cites increased public circumspection about the technology, noting a “rising movement of ordinary people from diverse professions, such as writers, artists, computer scientists, engineers, and philosophers, who found common ground against the (gen) AI paradigm.”
Andrei suspects that this rising tide of concern will only intensify in 2025 and suggests that many skeptics of AI feel “that it is being forced onto people by billionaires and their organizations.”
Increased Strain on Environmental Resources
Oxylabs’ CEO, Julius Černiauskas, shares Chaudhry’s concern that AI development is likely to be more scrutinized in 2025, adding that besides concerns about the safety of the emergent technology, there is also growing awareness of its environmental impact—particularly the energy usage that LLMs require. Just this month, Microsoft announced a partnership to reopen the Three Mile Island nuclear plant, a decision driven by the company’s increased investment in AI.
“Responsible AI and Green AI will become even bigger topics [in 2025],” Černiauskas concludes, “Servers that support AI development put a strain on the environment, and there are many risks flowing from a lack of transparency in how AI is developed and functions. AI companies must answer these concerns to earn and keep public trust in the benefits of this technology.”
Inflated Market Valuation
Adi Andrei concludes, “The Gen AI bubble will most likely burst in 2025.” Andrei quotes Tom Siebel, CEO of C3.ai, who has warned that “the market is vastly overvaluing it” and that “absolutely there’s a bubble.” According to Andrei, Siebel isn’t alone, and his views are common among CEOs and machine learning experts in Silicon Valley.
A Silver Lining?
Still, all three experts suggest that 2025 will also see exciting new technological developments. Chaudhry expects advances in multi-modal AI, especially text-to-video. Černiauskas anticipates that automated machine learning (AutoML) will open AI development to those who lack specialized machine learning expertise, effectively democratizing AI development and allowing specialists in varied fields to tailor machine learning to their needs. Černiauskas hopes that this will give companies more tangible reasons to invest in AI.
For Andrei, on the other hand, the exciting advancements in 2025 will not be based on AI but on a social paradigm shift. He expects that new information technologies will assist with a move toward more decentralized ways to think about life and community building and predicts seeing “most likely decentralized social networks, local currencies, and so on.”
Image Credit: Oxylabs
Rashan is a seasoned technology journalist and visionary leader serving as the Editor-in-Chief of DevX.com, a leading online publication focused on software development, programming languages, and emerging technologies. With his deep expertise in the tech industry and her passion for empowering developers, Rashan has transformed DevX.com into a vibrant hub of knowledge and innovation. Reach out to Rashan at [email protected]
