In November of 2022, ChatGPT went live and attracted 1 million users within its first five days of existence.

At the time, artificial intelligence wasn’t new, but suddenly, it was all anyone could talk about. Businesses began feverishly searching for ways AI could transform their organizations, and that energy and enthusiasm have not waned: Today, 97% of business owners believe ChatGPT will help their businesses. More than 60% of business owners believe AI will improve customer relationships, and roughly the same percentage believe AI will boost their productivity.

At Dualboot Partners, we have counseled many startups and seasoned enterprises on the potential for AI within their businesses. The process of consideration begins with three important questions.

Question #1. What do you want AI to do for your business?

Many people are attracted to the promise of AI. Per the stats above, it is celebrated for its potential to increase productivity, streamline operations and even improve customer relationships. But it isn’t always the right solution.

For instance, let’s pretend you run a business in which you have a team of analysts who produce reports. They do their job well, but you start to wonder if AI can do it better. In that case, it may be best to forgo AI in favor of your team. After all, you don’t need artificial intelligence when you already have actual intelligence. You know the exact output you will want, and your team knows the input they need to achieve their goals. Your operation is already a well-oiled machine, and there is no need to rock the boat.

There are, however, many instances in which AI can be tremendously helpful for a business, particularly when it comes to automating business processes, gaining insight through data analysis, and engaging with customers and employees. Understanding your goals for AI not only helps you build exactly what you need, it also helps you understand what you must have in place to make it work.

Question #2: What data do you have to train AI?

As human beings, we must go through a learning process to do most things in life. We have to practice walking before we can run. We have to learn code before we can program. Artificial intelligence operates in a similar way: It must be trained to perform the jobs we want it to do, and that training process requires a large amount of high-quality data.

Let’s put it in the context of the publicly available generative AI platforms such as ChatGPT or Bard: They operate by accessing the largest data collection ever assembled, the Internet; then they mine and analyze this data in seconds summarizing it information humans can consume.  The Internet is what informs these platforms. AI is simply a reflection of the data it finds around the web.

Now, you don’t necessarily need to have an Internet worth of data to power AI, but you do need a lot. And most companies lack the data set necessary to train their AI to perform its intended function. So before a company gets too far down the path of committing to AI, it is important to understand how much data you have – and how much more you may need.

Question #3: How much money do you want to spend?

This is a fundamental question for all software development, and it is just as much a factor in the development of artificial intelligence using a custom data set.

It is extremely expensive to train your own generative AI, and the cost scales with the amount of data you have to use. For an example of this, we can look to OpenAI, which created ChatGPT. They did not have the money to buy the computing power to do the work necessary to generate their artificial intelligence. So, to pay the bill, they partnered with a well-funded technology giant, Microsoft.

The future of AI, however, may not be so prohibitive. Many developers, including Dualboot Partners, are now combining custom, domain-specific data with publicly available generative AI platforms. Using this strategy, companies limit the amount of AI training required while gaining access to the comprehensive knowledge that underlies the Internet.

A few elements of note

Many questions and concerns have arisen over the quality of AI outputs from publicly available platforms, and it is important to remember that artificial intelligence is a reflection of the data it can access. If the data is flawed, the result will be flawed, so it is critical to ensure you are working with high-quality, trustworthy data from the beginning.

In addition, there are many claims that AI will eliminate jobs, much the way machines changed the shape of the manufacturing industry years ago. But we believe there is more opportunity here than cause for concern. And we are in good company: According to World Economic Forum research, AI is projected to create around 97 million new jobs.

At Dualboot Partners, we are excited about the potential of AI for the companies we serve and the broader market. We’re already seeing powerful results, and we expect much more to come.