AI and CES 2024: Will AI Continue to Dominate CES?
All images courtesy of the Consumer Technology Association (CTA).
Yes, AI will continue to dominate CES. AI is the new electricity. It will be embedded in most products, eventually becoming as expected and as non-differentiating as our power source of choice. But, like electricity, it will fade into the background as how it enables value comes to the fore.
Because AI was everywhere all at once, it dominated the dialog and distracted from the essence of CES, which, despite the admonition not to say Consumer Electronics Show from the CTA is a consumer electronics show.
People want to see, touch, and experience technology. Virtual reality faced its challenges because only those in the headsets received an experience, meaningful or otherwise. Those without headsets only observed people making what might, in other situations, be considered stupid gestures or acting in a way that might get them interviewed if they behaved in such a disconnected way in, for example, a diner.
AI doesn’t exist as hardware by itself. While there will be AI appliances, and it could be argued that an Amazon Echo is an AI appliance, the AI part of it is a service delivered through a device. The AI doesn’t show up in pictures of an Echo, like new contours or buttons do.
AI can exist in hardware, as some chips incorporate algorithms in silicon. But that experience goes like this: a chip manufacturer produces an image of a chip and says, “Our latest NPU! Look, a lot of neural processors.” That’s it. Until some application leverages the NPU for something interesting, it’s not much of a demo.
Besides robots and various analogs to robots, like intelligent drones, AI doing something is hard to photograph and touch.
What does AI in consumer products mean?
To keep it simple, I saw two types of AI at CES 2024. One was generative AI, or types of cognitive emulation, tools like ChatGPT that use a prompt to generate an image or text. At CES 2024 some of those examples were speech-enabled, which made for much more engaging demos. This type of AI seeks to mimic human communication and reasoning.
The WeHead avatar holding a conversation with attendees is an example of generative AI being deployed in associated hardware. GooVee, the game lighting company, shared that they are working on generative AI lighting scenes coded from a prompt. Typing something like, “Make my string lights flow like an Icelandic lava flow,” would result in a GooVee scene that translates images of Icelandic lava flows into a lighting experience. Baracoda’s Bmind Smart Mirror leverages generative AI to make health and wellness recommendations.
The Rabbit R1 proved a retro leap forward. It is a digital assistant that can look retro but uses an on-device action model to integrate apps on a phone. It seemed to me like an Apple or Google acquisition looking to happen, as its primary purpose is to integrate data across a phone ecosystem. I like the concept, but I’d like it on my phone rather than on another device. Unfortunately, for now, the R1 requires manual linking and doesn’t retain user information. To me, any device that doesn’t maintain encrypted credentials imposes an impediment to the tasks it professes to automate.
The other kind of AI is more subtle, and I think of it as autonomic AI. Rather than attempting a cognitive act, it performs adaptive functions based on “AI” models, usually machine learning. Autonomic AI reacts to situations and makes “intelligent” adjustments based on circumstances.
Examples of Autonomic AI include the Motion Pillow that detects snoring and adjusts its internal airbags until the snoring subsides. There is no conversation, just a sensor that detects a situation and then adjusts the pillow in the way its model suggests will correct that situation into a better state. Other products in this class include Invoxia’s and WIRobotics WIM™ (We Innovate Mobility) exercise assistant apparatus and the Nobi Smart Lamp with built-in fall detection.
At a slightly higher level than these products sit the autonomous vehicles, like the Yarbo Multi-purpose Intelligent Yard robot, which reacts to multiple inputs simultaneously but remains limited in its scope.
AI Does not need CES, but CES 2024 Benefited from AI
AI does not need CES. If AI works, it works in a distributed, data-driven world. There is nothing to touch or see or feel besides a momentary engagement. A smart mirror only looks smart when it offers advice, and that advice only applies to the person taking the advisement. Much of the advisory AI could just as easily come from a ChatGPT, Bard or Bing.
Televisions need to be seen. Headphones need to be heard. Refrigerators need to be bent into, their doors opened, and their drawers pulled out. Drones need to fly and buzz. VR headsets need to be worn with vests and gloves and controllers—without the hardware experience, VR isn’t an experience.
CES 2024 clearly benefited from the 2023 AI hype. But as much as it was everywhere all at once, you had to go looking for it to see it, and even then, attendees probably needed to have someone point and say, “See, there, that’s the AI.”
A couple of robotic-like items, including an updated WeHead, leveraged generative AI and made a good effort to make it interesting. Wehead demonstrated not only generative conversations running locally, but they did so with voice input and voice synthesis. I think they offered the most prescient demo at CES—the most useful AI will communicate with us naturally and do things we ask of it. The autonomic intelligences will continue to make their tiny adjustments, often unnoticed as many devices pick up that capability.
AI will likely remain a CES touchstone, but it will eventually manifest as just another feature. Perhaps the autonomic features will lose their AI association in favor of more organic descriptors.
Ultimately, AI will not have a pavilion or an area, as it will be so ubiquitous and embedded. While AI will remain everywhere all at once, future stories will focus on category value. Just airbrushing marketing materials with a false AI patina won’t be enough. AI will need to have an integrated purpose to rise to catch the eye of analysts, journalists, or consumers.
Will AI dominate CES in the future?
This year’s CES obsession with AI can’t continue. The AI-centric view must give way to solutions. It makes no sense to organize the show around AI as an abstraction. A hodgepodge of solutions ranging from automated lawn care or edge processing for the Internet of Things where the only shared feature is “intelligence.” As AI becomes ubiquitous, it will just be a feature within product categories. AI’s lack of specific hardware affords it no home at a consumer electronics show.
The story will be consumer electronics again. Many of them will include some form of intelligence, but if it doesn’t contribute to the value of the device within its category, it doesn’t matter. AI in lawn care, for instance, only matters in the lawn care industry—at least at CES. Sure, some underlying discovery may be repurposed for some other use, but that isn’t the job of CES. Those discussions happen at academic conferences.
2023 was one of those transition years where a new technology shaped the dialog. It wasn’t the first. I show my scenario planning clients an ad from 1932 that argues for the adoption of Electrolux gas-powered refrigerators. Natural gas, not electricity. The BBC declared 2016 the year AI would go from virtual to reality. That didn’t happen, but it sure looked like it might from the lines and exhibits at CES.
Another AI Winter?
As I sit in subfreezing temperatures this January 2024, I’m not suggesting that AI is heading toward another AI winter. I have seen a couple of those already. This round of AI innovation feels different, but feelings are not facts. AI will need to prove itself to designers and product managers—prove that it isn’t just a potential but a real contributor to value creation in the consumer space.
I have no doubt about generative AI’s contribution to businesses, but the use of generative AI in consumer products will require reflection. Our simple devices already recognize our voice, tell us about meetings, order groceries, and change streaming sources. The actions will become more sophisticated, but the interfaces will likely be similar. A voice interface won’t be a differentiator.
For AI to transform from novelty to companion, to avoid a new AI winter, developers will need to create applications that drive millions of people to use them repetitively, to integrate them into their lives, and perhaps even to trust their lives to them. Those models have yet to emerge, but we will know them when we see them.
The rise of the home operating system
Rather than a number of conversant devices, homes may benefit from an operating system (OS). That will be a new consumer experience. Apple, Samsung, and Microsoft already envision a home OS through their smart home efforts. Cable companies might also find this an appealing market. The rising dominance of Matter as the foundation for connected things will make smart devices more amenable to consumers.
CES may well have a home OS section in the future. The majority of the show will likely feature devices that integrate with the home OS. In 2025 and beyond, consumer electronics will still be about hardware, not software. AI will end up being table stakes. Devices will adapt locally and be scheduled, managed, and controlled centrally. A home will run security, manage software and firmware updates, and monitor connectivity and networking.
The full stack home will likely lead to homes with technology architectures of business IT. AI will probably play a role in design, monitoring, and the remediation of issues. AI has the potential to make home technology easier to adopt and use, but for now, that is only potential. The need to integrate various types of technology remains a future consideration, one with a poor historical track record.
Managing change
And as businesses know, technology changes all of the time. The promise of the smart, AI-enabled home will sound great until protocols get swapped, underlying hardware ages out, and companies decide to abandon features or product categories.
Everything looks great when it’s new and fresh. AI, as an abstraction, offers a lot of promise. As a long-term companion, however, it will likely face the same struggles as any other technology. Unique problems will arise because it will be so embedded—what happens when an AI becomes obsolete? Will new models run on old hardware? Will new knowledge representations be able to inform existing models? With the pace of change accelerating, consumers may need to face those questions soon.
CES will continue to show us the spectacle of the possible. Businesses and individuals will be left to manage the details of implementation and change management. The AI at CES 2024 offers a glimpse at what might be. Many factors will determine what actually happens. The themes of CES 2025 and 2026 will give us a window into how successful AI turned out to be and just how it manifested itself in the real world of the consumer.
A lot of uncertainty remains. Consumers need to demonstrate resilience as they face a future of AI companions at home that leverage the same technology that challenges their relevance at work. Consumer attitudes need to be monitored. It may well be social movements or legislative action that makes or breaks AI, not the underlying technology. The CES themes will act as a canary in the coal mine when it comes to anticipating AI’s future.
For more serious insights on AI, click here.
Did you enjoy this post on AI and CES 2024? If so, leave a comment, ask a question or like the post.
Leave a Reply