The term ‘Generative AI’ seems to be everywhere, and almost everyone seems to be using this tech already. But what does it mean, exactly, and how are Canadian companies using it? Read our survey analysis about how firms can use such tools without harming their business.
What we will cover
Artificial intelligence (AI) has witnessed tremendous growth since its inception in the 1950s. From text recognition and image recognition to even speech recognition, AI has developed into a compelling technology that has the potential to reshape various industries, from finance to healthcare, transportation, and manufacturing, to name a few. And the latest buzzword in AI is generative AI. But let’s first understand what generative AI is:
What is generative AI?
According to Gartner, ‘Generative AI refers to AI techniques that learn a representation of artifacts from data and use it to generate brand-new, unique artifacts that resemble but don’t repeat the original data.’
Simply put, generative AI technology uses massive amounts of unstructured data to feed its internal programming model to execute various tasks. For example, generative AI foundation models, Stable Diffusion and GPT-3 models, understand human language as written and spoken. In fact, GPT-3 enables its users to generate output in the form of text based on a text request while Stable Diffusion is used to produce realistic images when provided with a text prompt.
By helping businesses create new content with minimal human intervention, generative AI tools may present new opportunities for organizations to improve productivity, save time, and drive more revenue. In this context, companies should ideally define how they use generative AI tools and what strategies they should have in place to prevent misuse of such data.
To get a better understanding of how companies are utilizing the potential of generative artificial intelligence, GetApp surveyed over 600 full-time or part-time employees who’ve been using generative AI at least a few times per month. Scroll down to the bottom of this article for the full methodology.
How is generative AI different from AI?
AI technology is generally focused on identifying patterns, drawing predictions, automating tasks, and producing insights. On the other hand, generative AI tools usually focus on a prompt provided by the user to produce results or content as an outcome. While AI algorithms process data sets to produce outcomes, generative AI tools can generate new content based on training using available data.
The main difference between both resides in their abilities and usage. For example, traditional AI platforms are largely used for analyzing data and gathering insights while generative AI platforms are an advanced branch of AI that can be used to produce new data similar to its training data.
Generative AI tools are popular for content production and editing
Generative AI tools can potentially act as a transformative force in the realm of digital marketing and content production. They can do this by helping to automate content creation, improving customer interactions, and performing data analysis. According to Statista, the generative AI market value in Canada is expected to reach US$1.12 billion in 2023.
This comes as no surprise, given the rapid proliferation of large language models such as ChatGPT, a generative AI tool. ChatGPT reached over 100 million users in a time span of just two months, while it took other popular platforms —such as TikTok and Instagram— about nine months and two years to reach 100 million users, respectively.
To gauge usage, we asked survey-takers to select all the tasks for which they have been using generative AI tools, and text editing emerged as the most popular use case of generative AI. Nearly half (48%) of respondents use generative AI at work for text editing purposes, while 45% of them leverage it for text creation.
It is apparent from the data above that generative AI is typically being used for tasks that require a certain amount of creativity and effort to create various content types, including text, images, presentations, and analytics reports.
What are some different types of generative AI software that SMEs can use?
- Software for image creation: users can typically turn text into pictures and produce images based on a specific background location, topic, or style using some generative AI platforms. Such visual elements can be used for marketing purposes, among other uses.
- Software for coding suggestions: some generative AI tools can also help small and mid-sized enterprises (SMEs) involved in software development by providing coding suggestions and performing quality checks on the existing code sets. This can make it easier for developers to more efficiently maintain quality.
- Software for text-based applications: using some generative AI tools, businesses can produce original content by inputting prompts that can be used for publication via websites or other mediums. From essays to news articles, poetry, and blogs, some generative AI tools can help organizations generate a variety of content.
In the next section, we will discuss how generative AI is being used in the workplace by employees.
Nearly 8 in 10 respondents say generative AI tools increase their productivity
According to an expert at Gartner, “Generative AI provides new and disruptive opportunities to increase revenue, reduce costs, improve productivity and better manage risk.” Our study also shows that a combined total of 85% of respondents say that generative AI tools either significantly or slightly increased their productivity.
Moreover, around 7 in 10 respondents feel that this technology can help them save their company time.
When asked about which aspects of the work generative AI performs most effectively for survey respondents, creativity (46%) was found to be the most common aspect, along with the following other aspects:
- Innovation: 36%
- Problem-solving: 33%
- Analyzing and summarizing data: 33%
- Performing repetitive tasks: 27%
- Research data: 27%
- Inspiration: 26%
By enabling businesses to automate several time-consuming tasks, generative AI platforms can help free up more time to be creative. According to The Globe and Mail, Nusa Fain, director of the Masters of Management, Innovation and Entrepreneurship program at the Smith School of Business at Queen’s University, did an experiment on her students in which she gave them an hour to come up with a business idea without using technology. She then recommended her students to use ChatGPT to validate and expand the ideas they had come up with manually.
Within 15 minutes of using ChatGPT, they were able to explore the competitive markets and refine their ideas. This further adds emphasis on the point that generative AI technology can assist companies in saving time, improve business processes, and give them a competitive advantage.
What are the advantages of using generative AI tools?
While AI technology has traditionally been used to automate tasks, augment human decision making, and boost efficiency, generative AI tools may have the potential to transform more aspects of businesses. We investigated this by asking our survey respondents how generative AI is transforming their job.
Since some generative AI systems can take care of repetitive or time-consuming tasks, teams can dedicate more time to high-value or complex tasks that require human intervention.
ChatGPT is the most widely used generative AI tool
Now that we’ve discussed some of the benefits of using generative AI tools seen by respondents, the next question we wanted to explore is which tool is being used the most. ChatGPT is one of the most widely used tools for more than half of respondents (77%), followed by DeepMind’s Alpha Code (22%), and DALL-E (15%), among others.
What is ChatGPT?
ChatGPT is a natural language processing software driven by artificial intelligence that mimics human behaviour and enables users to perform automated tasks such as writing articles, drafting emails, and writing programming codes.
Such generative AI models are trained on massive amounts of information from the internet, including books, articles, or websites.
Of those respondents who admitted to using ChatGPT at work —whom we will refer to as ChatGPT users— these are the purposes they use it for:
Clearly, text editing is the most chosen response by ChatGPT users. When we asked survey-takers who use ChatGPT about the effectiveness of the results generated by the tool, a combined total of 98% said that they find results either highly effective or somewhat effective.
As a majority of ChatGPT users have found the results to be effective, such technology may gain even more traction in the coming years. Moreover, businesses may end up using such tools for a variety of creative as well as non-creative uses.
How often is ChatGPT used?
We’ve seen that ChatGPT is mostly considered effective by a majority of ChatGPT users. However, how often is it being used? Amongst ChatGPT users:
- A combined total of 51% use it a few times per day
- 31% of ChatGPT users use it a few times per week
- 15% use it a few times per month
- Only 3% use it less than once a month
Since only 3% of the ChatGPT users use it less than once a month, it indicates that respondents are using the tool fairly frequently. In this sense, companies should ideally have some regulations or guidelines in place to monitor the use of this new technology.
Let’s have a look at how survey-takers perceive the output of ChatGPT.
To learn if employees are concerned about the results generated by ChatGPT, we asked users to what extent they verify outcomes for errors:
- 49% of the respondents said that they perform some checks, but not for every output,
- 42% of them said that they meticulously review and verify every response before using it
- 7% of ChatGPT users spot-check the results
- Only 2% of respondents do not perform any verification on the results.
Since only 2% of this subset of respondents do not perform any checks on the output produced by ChatGPT, it may be inferred that such content requires verification. Users may not be able to rely completely on such content and some sort of spot-checking may need to be conducted as a rule. In that context, let’s look at some best practices to avoid getting inaccurate results and safely leverage the output of generative AI tools.
3 best practices to follow while using generative AI tools
AI-generated outcomes can potentially provide enormous advantages to businesses but may also lead to legal and regulatory risks. Below are some practices that organizations can ideally put in place to make sure generative AI tools are used cautiously.
1. Be specific and clear
When asking a query to a generative AI tool, try to be as specific as possible. Try to provide the required details and clear context that can help it understand the question better and produce more relevant results. For example, if you are looking to get the programming code for a video calling function with a specific feature, you should clearly mention your goal in the prompt, saying ‘write the code for video calling software that also has live chat functionality’.
2. Keep security aspects in mind
When using generative AI platforms for business purposes, security should be considered a vital factor. Even if your company has strong security policies in place, you might need a renewed strategy as generative AI tools can present new challenges. To reduce such risks, it may be important to have adequate security measures to protect confidential data from being used in queries submitted to such AI tools. If possible, organize security awareness sessions in your organization for employees on how to make the best use of generative AI tools safely.
3. Verify information from reliable sources
While such software attempts to produce relevant and accurate information, it may be ideal to verify such results/output carefully before using it in conjunction with your products and services. For example, if you’re seeking factual or crucial information related to research, cross-referencing that data with reliable sources can help ensure its accuracy.
In part two of this article series, we will take a look at what other concerns businesses may have surrounding the use of generative AI tools and what regulations they usually put in place to address those concerns.
Methodology
The data for GetApp’s Generative AI Tools Survey was collected in June 2023 and comprises answers from 660 respondents. We selected our survey sample based on the following selection criteria:
- Canadian resident
- Between 18 and 65 years of age
- Either full-time or part-time employed in a company and use laptops/computers to perform daily tasks at work
- Must understand the definition of generative AI: “Generative AI (GAI) refers to a type of artificial intelligence that is capable of generating new, original content such as images, videos, music, code, or text. It typically uses deep learning techniques and neural networks to analyze and learn from large datasets and uses this information to generate content that resembles human creations. Some examples of generative AI tools are ChatGPT, Bard, and DALL-E.”
- Must use generative AI tools at least a few times per month
