Generative AI: Shaping Business Trends for 2024

This year promises to be a turning point for the field of generative AI especially within the market research and consumer insights sector. Analytics and insights professionals should prepare for a digital landscape where Generative AI is seen not as an optional tool but as an unavoidable strategic ally. From the disruption of traditional business models to the rise of multi-modal outputs, AI has already begun making significant changes to the way we work.

Here are what we anticipate being the key trends that will define the AI landscape of 2024.

The rise of flexible business models

Insights companies, like consulting firms have traditionally relied on Time and Materials (T&M) business models – meaning a client pays for marked up time and resources spent on the project. However, the rise of specialized AI models is set to bring more disruption to the traditional system. We may see more firms choosing to shake up the industry by marketing AI for specialized tasks, challenging their conventional time and materials models.

[related-article title="Generative AI Ushers in a New Era for Insights: Knowledge Management 3.0" url="https://www.greenbook.org/insights/the-prompt-ai/generative-ai-ushers-in-a-new-era-for-insights-knowledge-management-30"]

Solutions like hybrid T&M and SaaS models, might offer clients the flexibility to leverage Generative AI capabilities while maintaining adaptable cost structures. This transition will be further boosted by the move from project-based to platform-based research products, allowing firms to challenge the status quo with AI-driven solutions for specialized research tasks.

Increased prompt size

In the context of generative models, like those in artificial intelligence, a prompt size refers to the length or complexity of the input given to the model to generate an output.

Expanding prompt sizes decreases the necessity for specialized models. With larger prompts in generative models, the demand for special-purpose models declines. When you make the instructions longer in these models, they get better at understanding and creating varied outputs in different areas. This makes them more useful for a wide range of market research applications, from more robust and in-depth qualitative analysis to multi step research methodologies or synergic qualitative and quantitative simultaneous analysis.

Bridging The Gap In Output Validation For Business Use

Output validation in AI is the process of assessing and verifying the results or predictions generated by an AI model. It is a crucial step in ensuring the reliability, accuracy, and effectiveness of the AI system's outputs.

There’s a growing emphasis on distinguishing between what machines can check on their own (automatically verifiable), and what still needs human confirmation (non-automatically verifiable), especially in fields like code generation. A "verified code" for business use that has undergone a thorough validation and testing process, that is robust, and error-free, can offer significant speed and cost benefits. By bridging the gap in output validation, the pace of InTech and analytics developments will be accelerated.

Transformed business communications

The significance of lengthy documents as a measure of effort or importance is dwindling. In our fast-paced industry, where time is precious, we’re moving towards shorter, more direct messages in corporate exchanges. The emphasis is increasingly placed on precision and clarity rather than extended text, reflecting a changing perception of document length as an indicator of value or dedication. There’s no doubt that AI tools will be helping in crafting these messages to maximise efficiency.

Simultaneously, there's been a surge in the use of multi-modal outputs in business communication - blending text with visual aids, video or voice-over elements, recognizing that extensive text may not always be suitable for efficient human consumption.

The rise of multi-modal formats created using AI, will potentially impact the way we deliver our results within the insights industry, potentially moving more into one pagers, video summaries or infographics as it offers a versatile mean to convey information, acknowledging the limitations of relying solely on text-heavy documents in modern business interactions.

AI's commercial landscape: quality, open-source trends, and margin challenges

A new trade-off in business is on the horizon, pitting the quality of Generative AI models against the price per token. "Price per token" refers to the cost associated with applying computer algorithms or techniques to individual units (tokens) in language-related tasks.

This balance will become a critical factor in favoring "good enough" models, finding the middle ground between quality and cost-effectiveness in various commercial applications. This shift will significantly sway purchasing decisions, as businesses prioritize models that strike an optimal balance between quality and affordability, redefining the AI market's dynamics.

Moreover, we’ll see the rise of open-source tools and models, compatible with commercial off-the-shelf (COTS) or mobile devices. AI applications will run efficiently on widely available hardware, making the technology more accessible and practical for a broader user base.

This trend, coupled with the quality-price trade-off, allows developers and researchers to collectively improve and advance AI technologies with lower or no costs. Particularly smaller organizations will benefit from advanced AI capabilities without a significant financial burden. The availability of open-source alternatives will contribute to a more accessible landscape for Generative AI applications, facilitating their integration into a broader array of systems and devices.

Meanwhile, as competition heightens and Generative AI services standardize, profit margins for per-token billing models are expected to shrink. Service providers must innovate to differentiate their offerings, focusing on value-added services and unique propositions. This trend levels the playing field for research firms of all sizes, emphasizing the quality and robustness of outputs over technology availability.

Data safety, ethical practices and regulation

Concerns over data safety are poised to diminish, especially among businesses with a genuine need for handling sensitive data, who may prefer on-premises processing for its practicality and security.

As AI's role expands in market research, upholding ethical principles to protect consumer data privacy and consent is paramount for maintaining trust. Transparency about how algorithms process and analyze data enables clients to trust the insights generated, boosting research credibility and compliance with global data protection laws. Anticipated regulatory frameworks will further delineate standards for data accuracy, privacy, and security, offering a competitive edge to firms committed to ethical practices and data protection.

New benefits of AI

The use of edge computing in market research enables the gathering and analysis of consumer data directly at the source, such as retail environments or through mobile devices. This allows for real-time insights into consumer behavior and preferences, providing businesses with immediate feedback to adjust strategies or enhance customer experiences. Not only does processing data locally reduces latency, but it also ensures privacy while eliminating the dependency for constant cloud connectivity, making it a valuable approach for time-critical market research projects.

By making a shift from automation to augmentation, collaborative AI emphasizes the synergistic potential of human-AI partnerships. By enhancing researchers' capabilities, AI enables us to analyze vast datasets more efficiently. This combination favors the quality of insights, as AI can identify patterns and trends beyond human capability, while researchers interpret these findings within the context of business strategies. Leading to more nuanced and actionable consumer insights.

So, this year we will see radical shifts in our sector, from the rise of flexible business models to larger prompts reducing the need for specialized AI models, making AI more versatile. The rise of "verified code" and multi-modal communication signals a shift towards automated standards.

Yet, to stay competitive, service providers must innovate. Meanwhile, on-premises data processing gains traction, addressing concerns about data safety. The trends in Generative AI for 2024 are poised to redefine how market research will operate and innovate in the digital age.