Moving beyond experimentation is key for businesses to use AI

Artificial intelligence (AI) is everywhere. In fact, since the launch of OpenAI’s ChatGPT 3.5 in November 2022, it’s impossible to escape it.

The benefits for businesses and other organisations have quickly become apparent, but it's important to get beyond the experimental stage.

According to IBM's Global AI Adoption Index, released in January of this year, approximately 42% of enterprise-scale companies surveyed reported that they are actively adopting AI into their business.

Additionally, 40 percent were found to be researching and experimenting with AI but not yet implementing any models.

Interestingly, new research published by the Reuters Institute at the University of Oxford suggests that heavy use of ChatGPT, the best-known generative AI, is still not fully integrated into many people's work practices.

The researchers found that “frequent use of ChatGPT is rare” after surveying users in Argentina, Denmark, France, Japan, the UK and the US. While only 1% of respondents in Japan used it daily, this rose to 2% in France and the UK, and 7% in the US.

“Of those who said they have used generative AI, most have only used it once or twice, and it is not yet part of people's everyday internet use,” the survey noted.

So which is it? Is generative AI the fastest growing technology ever, or is it still in its very early stages?

It's not just something that will happen in the future: generative AI is being used in business right now.

In truth, raw numbers may not tell the whole story, and those who work with AI every day understand its potential very well.

“I think it really depends on the nature of the work. For example, if you work outside all day, you're less likely to use AI,” said Peter Rose, group technology director and co-founder at TEKenable.

In fact, the adoption of generative AI varies widely not only across professions and industries, but also between organizations.

TEKenable has been using Microsoft's ChatGPT implementation to develop AI solutions for both itself and its clients and has received an overwhelmingly positive response. Users are leveraging generative AI for a variety of tasks that are quite different from approaches using other forms of machine intelligence such as machine learning (ML). One of the key benefits of generative AI is that its human-like interface makes it easy for users to engage with it.

Another benefit is that it can be used on unstructured data, meaning it can be used to help staff sift through reams of emails and internal company documents, as well as information formally captured in accounting, enterprise resource planning and customer relationship management software.

But if organizations want to get the most benefit, they need to prepare their data for AI.

“In customer surveys we've done across large enterprises and startups, 93% of executives agreed that data is critical — that means integrating data, looking at data quality, understanding your customer, and de-duplicating data,” Rose said. Still, GPT-4o could well be a new, transformative technology, not just an iteration, Rose said.

“I used it on two devices and set it up to essentially have a conversation with myself, and it did. The conversation was natural and flowed smoothly without any major interruptions. It's really amazing,” he said.

Importantly, OpenAI's technology, which is embedded in various Microsoft products including Azure OpenAI and Copilot, appears to have leapfrogged the competition. ChatGPT-4o was first demonstrated last month and quickly gained buzz as an improvement over its predecessor, GPT-4 Turbo.

“This technology doesn't just send audio off for processing and parse the text-to-speech results. It's much faster and more responsive than that,” he said.

Additionally, ChatGPT-4o is multimodal, meaning users can input any combination of text, voice, and images and receive responses in the same way.

This, combined with rapid response times, will make the technology more human and encourage organizations and individuals to experiment further with AI.

But Rose said that while experimentation is very welcome, it doesn't mean the technology can't be deployed today to solve real business problems.

“We all know AI is a breakthrough technology, but it's not just something that's going to happen in the future. Generative AI is actually being used in business today,” he said.