This article is part of our series exploring the business of artificial intelligence

Earlier this week, Business Insider reported that Amazon's Alexa Voice Assistant and smart speaker business is set to take a $10 billion loss. The news comes as competitors to Alexa grapple with their own difficulties and struggle to find a way to monetize the voice assistant.

The current state of Alexa and other Voice Assistants is a reminder of the gap between creating impressive technology and building a business around it. As tech companies scramble to lay off workers and prepare for the coming recession, cool but unprofitable technologies will have to find a way to demonstrate their worth or be destroyed to keep their organizations alive.

Eight years after Amazon's Alexa, here's what we can expect from the voice assistant's technology and business.

Good tech, but not good enough

Many innovations have enabled voice assistants like Amazon Alexa to do things that were impossible 10-15 years ago. Advances in automatic speech recognition help the Assistant pick up and interpret your voice under different background noise conditions, among other voices, and with different accents. Natural language processing systems powered by deep neural networks (morphers, RNNs, LSTMs, etc.) help assistants map different nuances of speech to corresponding commands (think of all the different ways you could ask about the weather or request a timer). And there are a bunch of application platforms, APIs, etc. that allow voice assistants to traverse the vast amount of information on the web and map voice commands to application functions.

The tasks are either so narrow that there isn't much room for error, or they are so insensitive that it doesn't do much damage if the assistant gets it wrong.

Once you try to perform tasks that are sensitive, require multiple interactions, or are multimodal in nature, voice assistants become unreliable. For example, consider making a purchase, one of the key use cases Amazon originally planned for Alexa. This is a sensitive task because it involves money and users want minimal errors. It's also a complex task because it often requires multiple steps, users want to see what they're buying, and they want to be able to browse through recommendations and alternatives. It's hard to do this with a voice-only interface. The same thing can happen with other tasks, such as scheduling meetings.

What do people pay?

So you've created a really cool voice assistant that does a bunch of tasks pretty accurately and others that are suboptimal. How do you monetize it? In the context of the current application, there are several solutions.

The first solution is to sell hardware, such as Amazon Echo or Apple HomePod or different variants of Google Nest smart speakers. In this case, business value will be tied to the price of the device, how many devices you sell, and how often customers replace devices. This applies to smartphones, as people spend hundreds of dollars every few years for an upgraded iPhone or Pixel phone. But it doesn't work with smart speakers. First, people are unwilling to pay a high premium for them because they don't use them very often. Second, smartphones don't have a whole lot to upgrade (basically mics and speakers, and sometimes displays). Therefore, customers have no reason to replace them on a regular basis. Finally, upgrading and maintaining the cloud services that power voice assistants also costs money. So basically, continuing to use smart speakers will increase costs for smart speaker suppliers, and ultimately, those costs will outweigh the profits made selling the speakers.

The second solution is to sell services. In this case, users pay a monthly or annual fee to use the voice assistant on their phone or smart speaker. In this case, your product must be valuable enough that users will be willing to pay for it. For this business model to be successful, your product must achieve product/market fit by solving an unsolved problem or creating enough added value in a use case to convince users to pay for it. Unfortunately, the kinds of tasks Amazon's Alexa and other voice assistants can perform aren't worth paying for.

Finally, you might think of Amazon Alexa as a channel to attract users to other money-making products. Amazon, for example, believes that Alexa will lead users to shop more online. But due to the limitations I mentioned earlier, Alexa doesn't provide a great shopping experience, and users still prefer to use mobile or web apps for purchases.

Basically, this brings me back to what I said at the beginning. From a scientific and engineering standpoint, Amazon Alexa is certainly impressive. But from the perspective of products and business, it does not have the elements of realization.

The next generation of voice assistants?

The first generation of voice assistants had a great idea (using your voice as an interface to your computer), but failed to create a profitable business model. We've seen this happen with VR headsets in the 1990s (too expensive and low quality) and AR glasses in the early 2010s (glasses that didn't provide enough added value to justify their price). Alexa and Siri still exist because they were created by very wealthy companies that, under normal market conditions, can afford to lose money on new products until they figure out the business model (or dispose of them).

What's next for voice assistants? I see some possible paths.

One solution is to wait for AI technology to get so good that it can power fundamentally different applications (e.g. voice assistant always one step ahead of you and reaching out to you instead of you calling it?).

Another solution is to move away from the current general-purpose voice assistant model to more specialized assistants that can be adapted for different applications. This will enable the integration of voice assistants into the context and workflow of applications and make them more capable of handling complex, multi-step tasks. I think this format could have a viable B2B business model, especially in industries that require a lot of hands-on work (manufacturing, restaurants, hotels, etc.), where voice assistants can increase efficiency and reduce costs. The added value for businesses will far outweigh the convenience consumers get from talking on their phones (much like Google Glass 2 found product/market fit in hands-on work).

I'm not sure that a no-display smart speaker will end up being the right form factor for future voice assistants. Many of our everyday tasks involve visual elements, and the use cases for voice-only assistants will be limited. But it will be interesting to see how it turns out.

Amazon has no plans to ditch Alexa just yet, but I think it might be time to move on to the next generation of the voice assistant.