I came across an excellent post on LinkedIn from Denis Panjuta.

In a great image, he brilliantly captured the different layers of the modern AI types and how they overlap with each other.

The big buzz today is around Generative AI (ChapGPT by Open AI, Bard by Google, etc.).  

What make Generative AI so appealing?  The simple explanation is that Generative AI can build any output (text, image, sound, or video) based on a giant model of sampled data (pre-training data).  AI before generative could only recognize, classify, and cluster, useful in many ways, but not applicable for generation of content.

If you really want a deep dive into how these models work and are constructed, here is a friendly Generative AI post that is technical, but still easy to understand.

Generative models have the ability to learn from data and create from what it has learned.  Furthermore, if you have a model that is trained in common language (like chatGPT), it can understand conversational requests.  This is what makes these models so usable for the average person.

For business, a conversational model can become a source of automation and efficiency if deployed correctly on top of an existing process.  This may sound easy, but it is extremely complex because training a model requires the correct data, elimination of bias, and sourcing the information. 

With that said, however, many professionals today are using tools like chatGPT to ask questions, check code, build process ideas, etc.  However, it is not a good tool if you want to create information based on your data alone.

At RevSparkAI, we have a platform that does exactly this.  We give you a working LLM model based on your data.  Once deployed, your model becomes your smart assistant that can create many of the manual tasks you currently have today.

Hope this was informative and if you want to see how we can help you, call us for a demo.

RevSparkAI Team