top of page
  • Writer's picturetransformcxdev

Exploring the Potential Applications of Chat GPT4

This is Tim Keefe with Transform CX, and welcome to another banter cast with my great friend Andreas Wieman from TTEC. We're here this afternoon to talk about chat GPT. 

We wanted to talk about our thoughts about it, some of the things we've seen in the past, and hopefully inspire others to think about how this excellent tool can be something we use going forward and various implementations.

The key takeaways from this banter cast:

  1. It's interesting to see the potential GPT has with everybody that interacts with it - from PhDs or likely PhDs and having them help them write research papers and people just playing around with it on helping write fictional stories. It's refreshing to see the interaction that the tool has.

  2. The reality is this is a tool and, like many tools. It's how you use it, where you use it. And what are you trying to do with it? Is that going to make a difference? And I think we're still out. We're still at, like, this is a shiny, sharp object. Let me play with it, and not cut a couple of fingers off in the process.

  3. It's more looking at a very pragmatic approach where we can apply GPT. If you have an agent, if you have to reach out to some of the brands to help with a technical issue that they're having with the device, let's say you're a wireless company. They have an iPhone 14 S, and we know that 15% of the customers that contact us have this specific problem with this iPhone device or the software related to the iOS release. As long as GPT has access to all of that data, we can feed it just like a child, train it, and give it access to the correct information you need to operationalize it. ChatGPT can help you get to those answers quicker. A lot of times, if you train that with an AI and pair it with Machine Learning, you can put that into messaging bot or a chatbot or something like that.

  4. Look at tools like GPT, natural language processing, and natural language understanding. They can process millions of lines of data very cheaply and very quickly. Out of the 100,000 calls you took yesterday, you will know that those calls were related to a product complaint or service complaint and start to operate with real data. That is a game changer because when we can begin to be knowledgeable, personal, and proactive with our customers or clients.

  5. Put Chat GPT in the hands of your data scientists and data analysts. It's removing the manual tasks they have to do. How things can get done is very practical. And then, we can measure it objectively. We could reach a point where we've stopped looking at things based on what happened. We can start to get to a point where we can be more predictive and model outcomes that we want to get based on aggregate data and not some minuscule sample size.

  6. One advantage of a tool like GPT is the ability for non-data scientists to start understanding what the data tells them. That will be powerful because we need to start looking at the outcomes we're trying to accomplish within whatever we're doing.

  7. I look at it in practical use to augment the current practice that's in place. And so that way, you're not putting too many eggs in one basket and one basket, and overextending yourself. That I have that expert in my ear, that expert that doesn't tell me what to do but can help me get to the correct answer more quickly. That's where GPT will be decisive when it can do these quick searches.

In conclusion, GPT can be a potent tool in the right hands. It can help to automate mundane tasks and provide insights into data that would take significant time to review manually.

With proper training and access to the right data sets, it has incredible potential for businesses of all sizes. As with any technology, it's essential to understand its implications and use it responsibly. With the right approach, GPT can help businesses make more informed decisions quickly and efficiently.

However, it is important to remember that GPT is still an emerging technology with limitations, so it should always be used in moderation.

1 view0 comments


bottom of page