CES 2025 was filled with real-world artificial intelligence innovations.
This year's technology fair was filled with unusual artificial intelligence products that featured questionable value propositions, serving as a showcase of what could be considered "low-quality artificial intelligence technology."
In 2025, companies still seem uncertain about the true useful applications of artificial intelligence. This perception became clear at the latest edition of CES, where kitchen devices, baby cribs, and other items were showcased that don’t seem to require AI. One example is Spicerr, a spice dispenser with a touchscreen that supposedly learns your preferences to suggest recipes. However, its usefulness is questioned since it does not grind spices and uses proprietary capsules that are not rechargeable, raising the question: was there really a demand for a salt and pepper grinder that suggests dishes?
Another highlighted product was Dreo’s ChefMaker 2, an AI-powered air fryer. This device can scan pages of recipe books and calculate cooking times and temperatures. Although its scanning feature sounds more rational than Spicerr's, the reality is that many consumers have not expressed interest in this capability.
However, CES showcased even stranger AI products. One of them was Razer’s Project Ava, a “gaming copilot” that plays in your place after capturing images of your screen and offering advice. This software has been controversial for being trained on gameplay guides without crediting their authors. Additionally, its functionality experiences lag and can interrupt the game audio, leading to questions about its actual need in the market.
The AI products presented at CES seem to reflect the overblown expectations within the industry. The previous year, AI companies raised $97 billion in the US alone, prompting vendors to experiment with ideas without facing many negative consequences, but with great potential for reward. The limitations of current AI have led to a trend of promising more than can be delivered. Applications like ChatGPT make mistakes, image generators are frequently inaccurate, and characters in AI videos often confuse with one another.
Thus, we find ourselves with uninspiring AI products: air fryers, spice dispensers, and gaming copilot systems. While they do not meet the true needs of most, they represent what is technically viable today with a relatively low investment in research and development. Next year is expected to bring more significant advancements.
