Engage for Conversation
Your AI-powered, enterprise mobile digital assistantDownload the Data Sheet
Generative artificial intelligence (AI) is at the forefront of everyone’s mind. Every day we hear about OpenAI, ChatGPT, the new Microsoft Bing search engine, Copilot, and other content-creating AI models. Of course, we all chuckled when we heard how it threatened to break up one reporter’s marriage after professing eternal and undying love for him. And how it claimed to want to steal the nuclear launch codes. But seriously, generative AI has very quickly entered and entrenched itself into the mainstream and has become impossible to ignore.
What is Generative AI?
Artificial intelligence (AI) is the science of building smart machines and programs that can perform tasks that typically require human intelligence. From self-driving cars and manufacturing robots to smart assistants and virtual agents, AI is becoming part of daily life, no matter your industry.
So generative AI is used to describe AI that can be used to autonomously create new text, images, video, audio, code, or synthetic data. Basically, any time it is generating something on its own, it can be referred to as generative AI. The term also includes learning algorithms that make predictions as well as those that can use prompts to write articles and paint pictures.
OpenAI’s ChatGPT is the single fastest-growing generative AI platform ever — it had 100 million users in just two months after launch. Even the trendiest of social platforms haven’t been able to capture that sort of growth. Instagram took two and a half years and TikTok about nine months to reach those levels.
Everyone is talking about OpenAI, which highlights the wide-ranging impacts that the platform is already having. For example:
- The FTC has published articles telling everyone to “calm down” with claims of AI enablement in applications
- Amazon has been dealing with the impacts of hundreds of thousands of AI-generated self-published children’s and young adult novels
The analysis of OpenAI has been filtered through just about every lens imaginable, including how generative AI is going to impact our society, our business, how we work, how we consume content, and how we communicate.
In this article, I’m going to look through a different, and hopefully more relevant, lens. Let’s explore how OpenAI is shaping the platforms we work with daily and leave ethical and other far-reaching concerns to the professionals.
The Business Impact
Clearly, the OpenAI platform is going to have broad-reaching effects on the type of work we do and will be felt across all technology sectors. The Microsoft platform will also be impacted. In fact, the integration of OpenAI into the Microsoft suite of products is already well underway. For example:
- Microsoft Viva Sales is evolving to become a sales-centered conversational platform that can make recommendations about the next email, message, or phone call you should send to a prospect or summarize and suggest action items based on information in your CRM. I will bet my own money on the premise that more than 90 percent of email traffic in the coming 18 months will be AIs responding to each other.
- Viva Insights has quietly become a powerful virtual meeting assistant — taking meeting notes, summarizing meetings, and identifying and assigning action items.
- The Microsoft Power Automate developer environment has been using the CODEX engine to make recommendations and, in some cases, write code. Microsoft hasn’t been shy about stating that “…anywhere in the ecosystem where a person would need to write code, they have the flexibility to start with natural language.”
These are all capabilities that are live and in production at this moment, and I think that even in the short term, we can see how there are many more opportunities to incorporate OpenAI into a host of additional Microsoft services.
Outcomes and Expectations
The real power of the OpenAI platform can start to be realized for organizations when we drive more specific and relevant knowledge sets into the data foundation on which the model is trained. So, one of the big challenges with OpenAI is the cost of training the model. As a result, it struggles with contextualizing current information.
Given the volume of data and the number of parameters, we estimate the model training costs somewhere around $4.5 million dollars and more than 15 days on over 10,000 servers. However, like everything in computing, these costs are bound to reduce quickly and dramatically.
If we think about the current capabilities of OpenAI models, they are very advanced. But the information they are trained on is publicly available and doesn’t contain information that might be specific to your business. Once we have the capability to augment OpenAI with information specific to your processes, it will be yet another example of transformative change we will have to navigate.
“80% of CX professionals believe AI will provide a better contact center experience and 79% believe AI will be more secure than interacting with human agents.” — Talkdesk Research
Here’s an example. By adding a model trained on business-specific information that incorporates customer support call logs, we can likely automate 80 to 90 percent of level 1 support calls and functions without any other net new technology. Think about it — AI trained on your specific resolutions, with the ability to communicate with users clearly and succinctly, and, maybe most importantly, with infinite patience. I think we can all see how this will have a transformative impact on the customer call center space!
We can easily see other near-term OpenAI capabilities being rolled out to the Microsoft suite of tools. While ChatGPT has been getting the bulk of the attention, we shouldn’t forget there are two other derivative models available in the platform — DALL-E, which can generate images and text based on a given input, and CODEX, which is designed to generate code from natural language.
I think one of the next integrations of OpenAI technologies with Microsoft will be to bring generated images from the DALL-E engine into almost all the marketing content platforms. Let’s be honest, designers and photographers are expensive and while stock photography has driven the price close to zero, the next logical step is to have DALL-E just generate my images for me.
This isn’t only to reduce the costs for photo and design services but to find just the right image is something that can be done simply by describing what you want to see. More importantly, we can adjust the images and copy of the message to be exactly what the customer is going to want to see. This ability to hyper-personalize content in near real-time is going to make the personalization efforts of today’s marketing and ad platforms look about as targeted as newspaper ads from the 1950s.
These are just a couple of examples that are imminent and are going to have an enormous impact. The reality is there is going to be an immense demand very soon for organizations that have knowledge of the OpenAI platform and the Microsoft suite of products.
I want to leave you with a parting thought about the potential power of the OpenAI platform, inspired by something I read recently from Nvidia CEO Jensen Huang. It is more of a question for you to ponder:
What if we are witnessing the largest democratization of technical capabilities ever? What if we realize human language is a perfectly fine computer programming language and that, with the advent of ChatGPT, computer programming is now accessible to everyone?
Hitachi Solutions and AI
With Hitachi Solutions’ combination of AI experience, industry expertise, and depth of technical knowledge across all the Microsoft products and platforms, I can’t think of an organization that is better positioned to take advantage of the advancements and investments that Microsoft and OpenAI are making in this space.
In fact, we were the first to market with a conversational AI solution in the Microsoft ecosystem. Hitachi Solutions Engage for Conversation is a seamless enterprise mobile framework designed to drive engagement and experience and improve enterprise efficiency by leveraging Microsoft Dynamics 365, Viva Sales, Power BI, Power Apps, Virtual Agent, Power Automate, and Teams. Bottom line — it provides conversational, chat-based AI in a simplified user experience.
One of our superpowers is helping you understand the Microsoft AI solutions landscape, determine which solution is best for your company, and explain why it is the best way to drive outcome-based value. If you would like to learn more about how Hitachi Solutions and Microsoft are empowering customers to achieve more with AI, watch my on-demand webinar: Innovate with AI and Cloud Scale Databases in Every App.
If you’d like to hear more from Hitachi Solutions AI and data and analytics thought leaders, be sure and check out:
- AI & The Human Element podcast
- How Do You Connect the Buzz Around OpenAI to a Secure Business Solution? podcast
- Engage for Conversation
About the Author
A self-proclaimed data nerd, unabashed technologist, and eternal optimist, Luke is chief technology officer at Hitachi Solutions North America.