Even when AI wasn’t the focus, it was everywhere at CES

AI was the big story at CES this year. AI was everywhere, from voice assistants powered by large language models to the rabbit R1. It was just a bit too much.

it was everywhere at CES :

This year may be the CES of AI, but these features have existed for a long time. It’s only that companies are now embracing artificial intelligence as a brand. AI is now a part of the mainstream. It’s hip and cool to put it at the forefront of a product. This shows that companies are forward-thinking and ambitious. This has led to the term being used wherever possible, even if it is not the AI that most people are familiar with.

How can we tell the difference between AI and other algorithms? What’s more, would this not lead to AI being over-promised?

The label AI is applied to anything, whether or not it uses generative AI. This gives the impression of something being new and exciting. Generative AI is also still in the throw-it-at-everything phase of growth. People are curious to see how far the technology can go and believe that it will make a difference. It’s why we see everything from Walmart restocking your pantry with AI models, to car companies adding ChatGPT to their dashboard so drivers have something to chat to.

Arun Chandrasekaran is an analyst with Gartner. He said that this is normal practice for many companies. However, it can lead to overpromising consumers if they discover that something marked AI isn’t really like ChatGPT.

Chandrasekaran explained that there is now a confusion between generative AI, and other AIs. This could confuse the field. “Marketers may be shooting themselves in their foot when they advertise a product that is not what people expect.”

Most people, for better or worse believe that AI is synonymous to generative AI, more specifically ChatGPT. This gives the impression that consumers expect a chatbot to “think” like a person if they use a product labeled as AI.

It is unfair to other products that also use AI. The robots that are roaming the CES floor, such as Samsung’s Ballie and LG’s AI agent robot thing, (which is not an AI agent, but AI agents are AI software which can perform tasks like booking a flight or finding a table in a restaurant) are adorable and engineering marvels. Their existence is more due to advances in robotics, and even computer vision, than the rise of LLMs. We don’t even know if Samsung used LLMs in order to train Ballie.

Then there is machine learning. AI experts will argue generative AI, and the foundational models that drive many versions of it, are just the next step in machine learning. No one talks about machine learning any more. Although it is considered “traditional” and old, I am sure that this technology powers many of the patterns recognition features on display at CES.

Technology has a lifecycle, and we may reach a point where people become disillusioned by AI’s promise, after it fails to solve the problems that they thought it would solve. Chandrasekaran said that this is when many innovations and more suitable use cases are born.

In the coming years, there will be features and products which do not require a chatbot. Just not at CES. Not yet.

Leave a Comment