We will no doubt look back on it as that. And we’re partnering with them. We’re involved, and that opportunity is coming our way. So long story short, we’re not €“ we’re going to stay out of the guidance business right now, but we’re going to try to disclose what we’re doing and the level of activity that we’re seeing.
Dana Buska: Okay. Can we anticipate it will be a growth year?
Jack Abuhoff: Yes, we’re very much focused on growth. So yes, it’s the easy answer to that.
Dana Buska: Okay. Excellent. Alright, thank you.
Operator: The next question comes from Marco Petroni with MG Capital Management. Please proceed.
Marco Petroni: Hi, Jack. How are you?
Jack Abuhoff: Marco, hi. How are you?
Marco Petroni: Good. A couple of questions. One, everybody has AI and machine learning algorithms, and they put them to different uses. But is there any company out there that you know of that combines that with the ability that you have on the data side with regards to organizing, collecting and overlaying synthetic data on top of that? Is there anybody out there, including the big guys, that can do that?
Jack Abuhoff: Well, I think there are. I think there are a couple of companies that are doing some things that are similar to us, though not very many. We’ve kind of got a view of the world that we can do two things well. And we think that there is like a virtuous circle that forms when we do the three things well. First is AI data preparation. We’re helping large companies accelerate their ability to innovate in AI by doing the things that we do on the data side. Second thing is we’re then helping deploy those models and integrate them into people’s businesses. So we’re helping build the models and then we’re helping integrate the models. And then thirdly, we have our own platforms, and we’re learning the hard way. We’re eating our own dog food, and we’re figuring out how to do it for ourselves first so we then develop the expertise to bring to both the data collection.
Well, how do you collect data in a way that results in high-performing models? And then on the model deployment, how do you best deploy models in legacy workflows and legacy systems? How €“ what are the opportunities for reinvention that you can bring to bear?
Marco Petroni: The ChatGPT it’s great to go in, but I’ve had experiences where I put the same data and it gave different answers. Obviously, that can’t be used within a company. Are you guys €“ do you guys have the same capabilities of OpenAI in terms of creating those type of ChatGPTs within an organization specifically for an organization, so for example, their call centers or internally to be able to use that to interact amongst employees as well as customers?
Jack Abuhoff: Yes. So the essential architecture behind GPT is an architecture that is also behind our proprietary Goldengate technology. Now do we have the same ability to stand up something that performs the way that, that one does in a generalized way? Absolutely not. We don’t have the budgets. Hundreds of millions of dollars was likely spent on getting it trained to the level that it’s been trained on. There is a tremendous amount of data that gets poured in to create the billions and hundreds of billions of parameters that drive model. A tremendous amount of cloud processing went into that. We cannot do that. But what we can do and what’s the future of the way these things will work is we can build on those. We can train them.
