Transparent AI Part 4: AI Applications

©2021 audEERING GmbH       Soroosh Mashal         03.03.2021

Here we are at the end of our series “Transparent AI” with episode four. This one will be about the applications of AI, types of AI, and the AI effect. If you have not seen the previous episodes, please make sure to check out part 1 and the following blogs on Emotion Modeling and AI Training.

Types of Jobs AI can do

Let us start with the applications of AI. In other words, it is time to find a job for our adult artificially intelligent agent in this world. There are two types of jobs we propose to our AI:

  1. Jobs that make our lives easier
  2. Novel Jobs that we as humans can not do

For example, an emotion recognition application in a call center can help the agents reflect on their tone objectively and adjust themselves in real-time if necessary. Besides, they can also see how the tone of the customer changes throughout a conversation and how much happiness they are bringing into this world. This example clearly shows an application of AI in which it helps humans shoulder to shoulder and augments them. One of our products, callAIser is designed for this purpose.

Health AI and Vocal Biomarkers

Another example are vocal biomarkers. Science shows that some diseases affect our vocal cords and our voice. These changes are so subtle that we can not hear them. However, an AI can look into the changes in frequencies and audio waves, and find patterns there that are otherwise hidden to us. It would then help us detect diseases like Parkinson, and hopefully COVID-19 from these changes. You can donate your voice to AI SoundLab and play a role in the creation of an health AI.

If we look at the gaming industry, we can see both of these categories there. entertAIn observe can help companies in their playtest process to ensure that their game resonates with their audience. In this example, AI saves you a lot of time to listen to every session and tells you which parts are more important for you to focus on. On the other hand, entertAIn play helps game designers and developers to create games and implement new emotional interactions that are otherwise not possible. In this case, AI is enabling us to implement completely novel scenarios.

Types of AI

Now that you know more about Artificial Intelligence, you are probably wondering about different types of it and the future of AI. In the past episodes, we talked about emotion recognition AI using audio which is a narrow use case. That is why we call it Artificial Narrow Intelligence (ANI). It can perform a task as well as a human or even better.

Most scientists think that by the end of this century, we will have artificial intelligence that can hypothetically learn and understand any intellectual task that a human being can. We usually refer to it as Artificial General Intelligence (AGI). It means you can possibly have an AI agent at work that can help you in a more humane way than searching in Google for hours to find an answer.

Scientists believe that at some point in the future of humanity, we will have Artificial Super Intelligence (ASI). ASI is a hypothetical agent that possesses intelligence far surpassing that of the brightest and most gifted human minds. In other words, the relation of ASI to our intelligence will be the relation of using our muscles to shovel and using a bucket-wheel excavator (BWE). It took mechanical engineers a few decades to go from a steam engine to a BWE, and now we are witnessing a similiar process for software engineers and AI researchers in their field.

The AI Effect

When looking at the evolution of AI and changes in the definition, we come across a concept called “the AI effect”. In simple words, it means whatever computers can do and we can understand, we do not call AI anymore. Back in the days, AI consisted of some algorithms that could understand their environment and interact with it. Now, it is a system that can learn and performs tasks that are considered doable only by humans. Maybe in the future, our expectations of AI would be to do super-human tasks and it would be completely normal for us.

Conclusion

So, in this series, we started a journey to understand the definition of emotion, and we learned how we can model them. Afterwards, we dived deeper into the world of AI, and we saw how an ANI can be trained using our data. Finally, we saw some of the applications that this “Emotion AI” can be used in and we learned some bonus points about different types of AI and the AI effect.

If you are interested in using our AI technology in your industry, please feel free to contact us. We would be glad to help you take one more step towards a brighter future.

Read more about audEERING