Enterprise (Multimodal) Generative AI: Paving the Way to Artificial General Intelligence

In the 17th edition of this newsletter entitled “Custom Generative AI: The Next Productivity Frontier,” it was concluded that the?“Customized Generative AI”?models could revolutionize both?“Enterprise Operational and Marketing Models”?and push the current productive levels to the next frontier by carefully selecting the?“Best-Fit Foundation Models”?that can turn the?“Enterprise Data Assets”?into?“Actionable Knowledge”?while carefully taking into consideration the relevant?“Critical Success Factors.”?Furthermore, it was discussed that the fundamental building blocks of these customized Generative AI are the?“Foundation Models.”?Simply put, the foundation models are?“Pre-trained open-source large language models”?that were trained on a massive amount of linguistic data sets and already consumed enormous resources to be trained in terms of infrastructure, time, and cost. These models can be fine-tuned using the?“Enterprise Data Assets”?to capture?“Enterprise Knowledge”?and be able to support the enterprise internally.

However, this edition of the newsletter will discuss the concept of foundation models and the approach to using them effectively from a broader perspective. The foundation models are, in general,?“Pre-trained open-source models”?that were trained on massive data sets and already consumed enormous resources to be trained in terms of infrastructure, time, and cost. What was eliminated from this definition and existed in the 17th edition are the words?“Large”?and?“Language.”?

The reason behind this elimination is that these categories of the models are not usually limited to the language domain, and their size is not always large. Currently, there are many foundation models that are specifically trained on images to generate images, and some are trained on voice data to generate voice, and so on. Furthermore, although the actual value of the generative AI models started to get noticed when the large models dominated the market, research and development efforts are currently focused on producing the same impact using smaller and smaller foundation models.

This directs the discussion towards a new productivity frontier. Imagine what can happen if these different foundation models are combined into a single AI system that can react to text, image, and sound inputs and generate similarly. This is what can be called the?“Multimodal Generative AI Systems.”?Going a step further, if these models are designed to empower the robots, building a robot that can read, listen, see, and then respond with voice and actions similar to humans will be possible. Many experts believe that these multimodel generative AI systems are opening the way towards what is called?“Artificial General intelligence.”?

With artificial general intelligence, human intelligence can be simulated totally. Although this turning point hasn’t been reached yet, we are heading toward this point at an accelerated pace.

The second eliminated word mentioned above was the word?“Large.”?Currently, and after understanding the exact nature of the language models and how these models consume energy, computational power, and time to be trained as well as producing a tremendous carbon footprint during daily operation, the research and development efforts focused on creating the smaller size of the foundation models that can outperform the large models in many use cases while reducing the needed energy, computational power, and time to be trained as well as reducing the carbon footprint during the daily operation.

Hence, and to conclude, the?“Customized Generative AI”?models can push the current productivity levels to the next frontier by carefully selecting the?“Best-Fit Foundation Models”?that can turn the?“Enterprise Data Assets”?into?“Actionable Knowledge.”?More importantly, several types of these foundation models can be combined to form what is called the?“Multimodal Generative AI Systems.”?Furthermore, these multimodal systems can be built from the?“Small and Specialized Foundation Models”?that provide better total cost of ownership and less harm to the environment. These two trends in generative AI research and development are paving the way to?“Artificial General Intelligence”?at an accelerated pace.