For the past 18 months, generative AI has taken our world by storm. From writing poetry to doing our homework, it seems like that it can do everything.
Today, no innovation can be discussed without considering either how AI will enhance it or make it redundant. Generative AI offers profound transformations across industries.
But it’s not without risks. In this post, I’ll outline a few use case categories for innovating with generative AI and address some risks with intellectual property (IP), privacy, and hallucinations.
1. Enterprise Technology Development
This was one of the most surprising fall outs of generative AI.
The applications of AI are not beyond simple acceleration of development processes. Gen AI is set to redefine this space. Not only can robot developers reduce time and enhance consistency, but the low and no-code platforms must redefine what they do. Coderbotics has set out to do so, as have many others.
But, there is a silver lining.
The pace of innovation will rapidly increase once we move past the “storming” phase of the AI-human teamwork. At that point, the real value will be to ensure that software iterations align well with the overarching enterprise goals. Instead of competing with AI, we’ll achieve more by working in tandem.
As a result, not just technology architecture, but business architecture will become key skills to have. It’s reasonable to assume for the foreseeable future that humans can beat AI in figuring out what will meet a market’s need given all the nuances of nature.
2. Insights Generation
Generative AI is making great strides in analytics.
For example, in marketing, generative AI takes Customer Data Platforms (CDPs) to the next level. It enables marketers to discern nuanced patterns. This in turn facilitates more informed decision-making.
Beyond mere content creation, generative AI can craft personalized narratives that resonate with individual customers.
In fact, as the introduction of gen AI in MS PowerBI shows, asking for insights will never be the same again.
Business Process Improvements
Let’s consider these with examples from banking and insurance.
In the insurance sector, generative AI can transform claims processing. By extracting intricate policy details (see also the risks section below), it simplifies eligibility checks, coverage assessments, and payment procedures.
It also works well in handling complex policies. Insurance claims, once a labyrinth of paperwork, may become a pleasant experience.
As another example, generative AI’s impact on consumer banking is that it can evolve into a financial buddy that can dynamically analyze credit transactions and offer tailored advice.
This not only expedite processes, it personalizes the banking experience.
Customer Service Transformation
The difference here is actually pretty stark.
Almost everyone dreaded the chatbots till last year. But now we are beyond the scripted responses of traditional chatbots. Gen AI driven chatbots comprehend context and navigate complex conversations easily. They can interpreting user intent and tailoring responses based on the latest content of the website and any other data sources given to them.
This introduces a new era of customer interaction, where support becomes not just efficient but remarkably personalized. In fact, this technology can also significantly aid human interactions.
Risks
Generative AI is not without risks. Here are 3 areas to be cognizant of as we think of use cases to implement.
Intellectual Property
As AI generates derivative works, the potential for unintentional IP infringements arises. We need to tread carefully, ensuring that innovative pursuits do not step on existing IP boundaries. The jury is literally still out on this issue so we’ll have to wait and watch.
Privacy Safeguards
Privacy is a key consideration in the deployment of Generative AI. Providers like OpenAI recognize this imperative and are offering robust privacy guarantees that enterprise inputs being received will not be used for training.
Hallucinations
However great the technology is, the core of it is still based on predictions, not knowledge and judgement. So the output can look convincing and still be entirely wrong. Of course that can be said of humans too, so whatever checks we put in place for humans should be in place for AI too.
Next Steps
Generative AI’s impact on speed of innovation can be significantly enhanced with automation and integration.
So we should start thinking in terms of how to build business ecosystems to cement our competitive edge.
That will require a mindset change away from storming with AI to norming and performing with AI.