Guest IDC Blogger: Matt Arcaro

As we approach the halfway point of 2024, IDC continues to see the increasing role of AI in both our personal and work technology personas. In many instances, AI will merely remain a technology enabler powering a capability or feature. More importantly, in others, AI will become deeply entrenched in the product development and engineering life cycle, delivering new experiences and improving user engagement, visibility and productivity. However, the road leading to this AI-driven future is not straight and predictable but rather winding, including frequent stops, starts and failures. Avoiding these AI adoption and integration pitfalls is critical, so IDC offers insights into five of the more prominent AI myths that arise in discussions with technology buyers, suppliers and the technology ecosystem at large

Myth 1: Organizations have already adopted AI en masse.

Reality: Although AI has become a key part of how everyone talks about the progression of today’s technology, the reality is that only 59.3% of 361 global CIO and IT leaders in IDC’s “Worldwide CIO Sentiment Survey” (October 2023) indicated that they have adopted and deployed AI in their environment. Although this may be the majority of organizations, a large continuum remains in one of the technology pre-deployment phases of research, experimentation, procurement/vendor selection, building, testing, prototyping or validation of AI solutions. IDC views this current mixed state of AI adoption as a reminder for technology buyers and suppliers that deploying AI is not a switch to turn on but a multi-pronged effort spanning technology, processes and people.

Myth 2: All AI initiatives deliver strong ROIs.

Reality: AI has delivered transformative capabilities, features, experiences and insights to a range of organizations. Yet organizations, especially those new to AI adoption, have not seen the quick payback results they may have hoped for. For example, in IDC’s “WW Future of Operations Survey” (August 2023), only 28.7% of the 864 respondents indicated that their AI data initiatives realized a rapid payback. This is to be expected and highlights what technology suppliers and buyers often forget: AI is not a panacea but a purpose-built tool with strengths and weaknesses. Technology suppliers and buyers must prioritize strong governance and use cases to determine where and when to position and utilize AI and establish metrics to ensure that it is delivering measurable value.

Myth 3: Traditional AI is going away. GenAI is all that matters.

Reality: There has been a wave of technology suppliers and buyers asking us about GenAI in the last year. The LLM advancements and multimodal models largely driving GenAI growth have forced organizations to rethink how and where they can utilize AI, especially for unstructured data tasks — but this isn’t the entire story. Many of the advancements underpinning GenAI have led to improvements in the use and applicability of traditional, predictive, and interpretative AI. In IDC’s “Worldwide GenAI Arc Survey” (September 2023)of 1,304 technology decision makers, respondents indicated that they expected 80% of their overall AI investment budget to be allocated to traditional AI initiatives versus GenAI. This makes sense, as traditional AI, although less talked about today, is better understood and often aligns better with organizational edicts to move the needle via productivity, visibility and data insight improvements.

Myth 4: Organizations must always build AI solutions themselves.

Reality: There are a range of benefits for companies in building or customizing technology solutions with AI. These developed and deployed capabilities may deliver something unique that can act as a moat to defend against existing and emerging competitors. The challenge to this approach (as referenced previously) is that building and deploying AI is not without risk and can be costly, time-consuming and unsuccessful. IDC suggests that in many cases, especially for IT and operations functionality, a more practical approach would be for organizations to leverage AI capabilities included within their supplier’s technology and product offerings. This serves two purposes: it reduces the level of effort required to bring AI-powered features into an organization and it facilitates and builds internal experience and alignment to better understand the ins and outs of AI.

Myth 5: AI is AI. The technology supplier building the AI-powered capability doesn’t matter as much as it used to.

Reality: This is one of the more interesting myths that we a lot about. Although the high-level approaches of competing vendors may integrate and utilize AI in similar ways, IDC argues that a vendor’s approach matters more now than ever. This is driven by a few different principles, most critically the amount, relevancy, and diversity of data a vendor uses to build a given AI-powered feature. Remember, not all data is created equal, and technology suppliers may lean into AI models’ “black box” aspect to deploy capabilities that may work but are not (defensibly) enterprise-grade. In other words, IDC tends to see more success stories when technology suppliers build AI-powered capabilities into their products and services that focus on solving known customer and/or business problems. These suppliers utilize an outcome-first lens to build and extend customer trust through AI that leverages a progressive, use case-by-use case strategy.

IDC strongly believes in the power of AI, but much like any new technology, the market’s hype often obfuscates some of the nuances and realities as its use continues to grow. Hopefully, the five AI myths and realities highlighted above help you better understand the current state of AI and enable you to navigate AI’s complexities more effectively to drive tangible business outcomes.

The post Setting the Record Straight on Five Key AI Myths appeared first on IT Glue.