ABI Research took an unusual approach to predictions for 2025, highlighting a range of technologies and trends it does not expect to make it big rather than offering a glowing report of likely success stories.

AI RAN and smart glasses were among the 101 technology trends the research company listed. On the former, it explained although while SoftBank completed a trial in 2024, it does not expect actual deployments until 2026, with broader adoption likely in 2027.

Chief research officer Stuart Carlaw explained ABI Research expects “trials and pilot projects will advance the technology” in 2025, “but large-scale deployment will take longer to materialise”.

In a whitepaper, ABI Research noted “many industries and end markets are in that awkward stage of technology adoption”, as they ponder implementation strategies, assess technologies and partners, and decide if they have the resources to deploy services and applications at scale.

“This is a particularly sensitive time, which tends to suggest 2025 will have tech implementers and end users on the brink of a period of a massive technology shift as they work through these issues.”

The company believes the consumer smart glasses sector will also not see widescale adoption.

“Despite the significant strides we will see in AI, smart glasses are expected to remain predominantly in the realm of early adopters”.

The company expects 2025 to be another transitional year for smart glasses, “with most traction coming from niche applications rather than widespread consumer or at-scale deployment”.

On the plus side, VR is tipped to maintain a strong grip across gaming, sports and fitness, sectors which have driven consumer demand.

ABI Research also popped the bubble on semiconductor onshoring efforts, arguing 2025 would be “a pivotal reality check for US and European” efforts “as implementation challenges force significant delays”.

It noted despite massive funding for US-based chip manufacturing through the CHIPS and Science Act, TSMC, and Intel‘s domestic and European ambitions “face setbacks due to EUV machine shortages, workforce gaps and operational complexities”.

“Full operational capacity is unlikely before 2027, jeopardising the EU’s goal of 20 per cent global chip production by 2030”.

Instead, obstacles will lead to strategic reassessments in 2025 and shift the focus from full onshoring to “pragmatic” diversification and trusted partnerships.

“The semiconductor cold war may also slow innovation in AI and other advanced technologies reliant on cutting-edge chips”.