The Rabbit R1 is a perfect story for the moment. CES 2024 was heavily focused on Ai, but it has provided a great time to talk about the voice assistant and hardware in general. For users with low to no vision, voice-based assistants are a valuable tool. For everyone else, it’s questionable.
There have been many recent announcements in Ai and they are mostly centered around generating images based on text prompts or responding to text prompts like a chatbot. For those who didn’t explore AOL instant messenger, the chatbot may seem entirely novel.
Dedicated chatbot hardware? That’s a fun idea, but here’s why we think this is a difficult proposition.
Device Collapse
The smartphone has devoured an array of gadgets. The original iPhone wanted to be the iPod, phone, and Internet device. Since then, it’s also distrupted the still camera, video camera, TV remote, navigation, and perhaps even the menu at a restaurant. A modern phone wants to be everything.
Walled Gardens
The iPhone has Siri. Samsung has Bixby. Other Android devices offer direct access to Google Assistant. Echo devices get Alexa. These options are isolated from each other. The wake words are mostly baked into whichever platform you select. There are few exceptions.
Money Problems
Voice assistants haven’t been lucrative so far. Beyond the hardware cost, it appears that people are not willing to pay a subscription for this type of product. But Humane’s Ai Pin and Rabbit R1 are betting the big tech companies are wrong about voice.
Is Ai the secret sauce for a compelling and lucrative voice assistant? Do people want to carry a second device with a service plan to escape the ones their phones require? What do people want to do with an Ai-based voice assistant? Is it impolite to converse with these devices in public instead of the other humans nearby?
The bigger question is about financial returns. The dominant players in the traditional voice assistant category are owned by Google, Amazon, and Apple. These companies technically don’t need these products to turn profit. Most other companies will probably need to charge monthly to demonstrate their market growth potential. It’s gotta be SaaS!
How much are companies spending to train and infer from these Ai-based models? That will determine how much they’ll need to charge monthly and if or when they’ll break even. So far, experts can only speculate about how much these companies are spending on the Ai future. Despite being called OpenAI, we don’t know how much it costs to run.
These companies have a critical interest in projecting feasibility and hooking users. It only works when people know and use the features. People will only use Ai if they already know it can save them time and generate grounded answers. Companies will only continue to invest in Ai when it proves to be profitable in the long term. That’s why it’s important to separate from the hype and evaluate these solutions based on their merit and value rather than our hopes and dreams.