Technologies that change the way people interact with the world.

Technologies that affect operations through changing practices, processes, or functions.

The joining of multiple trends and technologies that help organizations to accurately classify, predict, and solve a large volume of problems that are difficult for humans to do in a short amount of time.

Advanced virtual assistants are also known as AI conversational agents. They process human inputs and come up with predictions and decisions. AI conversational agents are driven by a combination of conversational user interface, natural language processing (NLP), and semantic and deep learning techniques which include the following:

• deep neural networks (DNNs)

• prediction models

• decision support

• personalization

Time to market: 1-3 years

The estimated time to market depends on how quickly the current virtual assistants change over to advanced virtual assistants. This then can expand into all areas of consumer lives, business interactions, and operations.

Advanced virtual assistants have a high potential impact because the technology can be used in almost all disciplines. It has the potential to transform how an application is utilized and how users interact with devices while enhancing their experience and engagement. For example, there is AI in AT (more specifically AAC) to help text-prediction so that the user can communicate at a faster rate.

Transformer-based language models are deep neural networks (DNNs). In this model, words are processed as sequences in a sentence. This keeps the meaning of surrounding terms so that it can improve translation, transcription, and generation of natural language. These models are trained on large datasets of billions of phrases.

Time to market: 1-3 years

The time to market depends on how effective the training tools are and the ease of usage. Transformer-based language models can write paragraphs that are similar to those written by a well-educated human being.

Transformer-based language models have high potential impact because they are quickly taking over the place of recurrent neural networks (RNNs) systems. New tools are delivering great improvements in advanced text analytics and related applications (i.e. automated text generation and intelligent virtual assistants).