Generative AI
Generative artificial intelligence (AI) is changing the way we interact with technology, create content, and run our businesses at a pace not seen since the early days of the Internet. Conversation about this exciting technology is everywhere, and interest in its potential across industries has been dominating headlines, online webinars, and corporate events for some time now.
A recent Gartner survey revealed that business leaders believe that generative AI is he technology that will have the greatest impact on your business in the next three years. In this context, it’s easy to feel compelled to dive headfirst into the world of generative AI, but for small and medium-sized businesses (SMBs) with limited time, budget, and resources, it’s rarely the best course of action.
Ben Schreiner, director of business innovation for SMBs at AWS, says that, like any technology investment, “the decision to implement generative AI should be the result of a thorough analysis of the return on investment (ROI), supported by a case of well-defined business. Before you can do that, it’s important to understand exactly what generative AI is and how it can help SMBs.
What is generative AI?
Generative AI creates new content based on cues provided by users. Applications powered by generative AI are capable of holding conversations and answering questions, as well as creating content such as articles, videos, images and audio. Whatever the use case, the generated content is very similar to what a human being could produce, only it is produced in a much shorter space of time. Think in minutes, not hours or days.
All generative AI applications are based on basic models (FM) that are trained with enormous amounts of data. Different types of models are trained for different purposes. For example, large language models (LLMs) can be used to generate text, answer questions, and help developers code, while fuzzy FMs are used to create and edit images, videos, as well as 2D and 3D models.
How can SMEs use generative AI?
The main benefit that SMEs can obtain from using generative AI is efficiency. According to a recent Prosperous Insights and Analysis According to the survey, the majority of SMBs currently using generative AI do so to assist in research (40.3%), with other popular uses including content creation (29.6%) and customer service (26. ,2%). While these use cases are impressive, they only scratch the surface of what generative AI is capable of.
Prosper: what ChatGPT is used for
Generative AI can be used in unique ways across different industries. Financial institutions can use it to detect fraud and improve customer service with chatbots. Healthcare companies can simulate clinical trials based on patient data and accelerate research from years to weeks. Manufacturers can generate new part designs and automate inspections. In the world of media and entertainment, companies can automate tedious translations into other languages and more.
The importance of working backwards
For SMBs with a finite budget and a few free hours, it’s important to avoid the temptation to implement generative AI prematurely. As Schreiner explains, “Working backwards from a significant challenge to the business or customer and performing a thorough ROI assessment is crucial to making an informed decision. “That way, SMBs can be confident that the potential returns from generative AI justify their investment.”
He continues: “Once a business case has been established, SMBs can leverage pre-trained generative AI applications, such as chatbots and coding buddies, or create their own. It is important to note that the latter may require some training or collaboration with an external provider depending on the internal skills available.”
An AI solution is only as good as the data it is trained on
One of the main strengths of generative AI FMs is that they can be customized with a company’s own data. Schreiner explains: “SMBs should consider securely storing their data in the cloud (the same cloud as the FM they intend to use) to make this process as easy and secure as possible.”
Many small businesses struggle to separate good quality data from bad and suffer from fragmented and isolated data sources. Adopting a cloud data strategy, supported by modern data architectures such as data lakes (centralized, scalable, flexible and cost-effective data repository), can remedy this problem and leave SMBs in a strong position to implement the Generative AI.
The choice of model (both type and size) is also important when creating generative AI applications. “The amount of data needed to achieve a given use case will vary, and SMBs should avoid choosing an FM that far exceeds the amount of data needed to achieve their goals,” explains Schreiner. “Experimenting with different FMs before committing to one is key to remaining profitable.”
Responsible AI
Generative AI is new territory for both companies and policymakers. Schreiner notes, “There are multiple dimensions of responsible AI use that SMBs should consider, including explainability, fairness, governance, privacy, security, robustness, and transparency.”
The legal parameters around the use of generative AI are evolving rapidly. Staying up to date with the latest laws on copyright and intellectual property (IP) infringement is essential for any company using generative AI. SMBs should avoid sharing content that closely resembles that of other companies or creators, while also taking steps to prevent their own data from being used to train other FMs.
“SMBs should also think about algorithmic fairness related to diverse and inclusive representation and detection of bias,” explains Schreiner. “Guardrails allow SMEs to define a set of disallowed topics that are undesirable within the context of their application. This allows FMs to automatically detect and filter unwanted content in training data, input cues and generated outputs.”
While generative AI can create persuasive content, it can’t always be trusted as a reliable source of information. As with any content published by an SME, it is advisable to independently verify the accuracy of what is said or shown before sharing it publicly.
Most SMBs aren’t using generative AI…yet
While it may be tempting to start using generative AI as quickly as possible, fear of missing out (FOMO) is not a justifiable business case. But it is true, however, that those who actively (and carefully) consider implementing generative AI are ahead of the game.
Going back to the Prosperous Insights and Analysis According to the survey, 18.7% of SMEs have never heard of ChatGPT and 32% have heard of it but don’t know what it is. It’s important for SMB owners and decision makers to remember these numbers whenever they feel pressured to hastily implement generative AI.
Prosper: Have you heard of ChatGPT?
Bottom line: Generative AI is redefining what is possible for SMBs, and the decision to implement it will likely be one of, if not the most important decision small business owners will make in the coming decades; so why rush?
Working backwards from a significant challenge, conducting a thorough ROI assessment, ensuring data readiness, and establishing clear policies for the responsible use of AI are essential for SMBs looking to make generative AI a success in their business. business.