6 Reasons Why Companies Still Hesitate to Embrace AI

Welcome to DX Weekly — a free newsletter to help you unleash the power of digital for your company. Subscribe here. Digital transformation is the antidote to any company becoming irrelevant, and each week, I’ll share new actionable information to help you and your organization succeed in the digital world.

AI adoption is on the rise, but many companies are still unsure about implementing it fully. A 2023 survey revealed that while 76% of companies used AI in some capacity, only 9% widely adopted the technology.

Are you one of those struggling to get your organization on board? Why is it so hard to get big companies to adopt AI?

These are the six main barriers organizations face when trying to implement AI initiatives:

1. AI HALLUCINATION

A major concern with generative AI or computer vision tools is their potential to generate inaccurate, fabricated, or nonsensical outputs, known as hallucinations.?

Hallucination can stem from bad training data or misinterpretation of the data. While AI is good at finding patterns, it sometimes detects patterns that are not there, similar to how humans might see figures in the clouds.?

One way to mitigate hallucination risks is to use the right AI tools for your needs. General-use chatbots like ChatGPT can be good enough for less sensitive tasks like writing generic emails or coming up with content ideas. However, for high-impact business use cases, domain-specific models or AI tools trained on your own knowledge base can provide more relevant and reliable information.

You can also minimize hallucination by creating clear, specific prompts that provide adequate context and parameters. For example, instead of saying, "Tell me about climate change,"

say: "Explain three major effects of climate change on global sea levels, use data from the last 20 years from reputable scientific organizations.”

Another approach is to adjust the chatbot’s “temperature”, a parameter that controls the creativity of its responses. A low temperature yields a more focused and “safe” answer, while higher temperatures increase creativity but can lead to incoherent or irrelevant outputs. Developers set default temperature settings for chatbots, but some models allow adjustments on the back end.

Finally, human oversight remains crucial in ensuring hallucinations are caught and corrected before they cause serious issues.

2. KNOWLEDGE GAP

According to research, 37% of organizations hesitate to adopt AI because they don’t really understand how it works.?

The complex nature of AI models and their underlying algorithms can make it challenging to effectively communicate their functionality and potential impact to stakeholders who don’t have a technical background.?

Without a clear articulation of its business value, decision-makers may struggle to justify allocating resources to AI initiatives.

To overcome this barrier, you’ll need to educate yourself and your team about the practical applications of AI in your industry. Seek real-world examples and consult with AI experts to better understand how AI can solve business problems and drive results. AI consultants can help your organization understand the potential benefits of AI and make it easier to get investors on board.

3. COST

The cost of developing and implementing AI solutions can be substantial, ranging from tens of thousands to millions of dollars, depending on the scale and complexity of the project. These upfront costs can significantly deter companies, especially without a clear understanding of the long-term return on investment.

While the initial costs may seem intimidating, successful AI implementations can lead to significant long-term cost savings, increased efficiency, and improved business outcomes.?

To navigate the financial challenges, start with focused, high-impact AI projects that can demonstrate quick wins and provide a foundation for future investments. Partner with experienced AI implementation consultants who can help optimize costs, maximize ROI, and continuously measure and communicate the value generated by AI initiatives.

4. SKILL DEFICIT

The shortage of AI talent is another significant barrier to AI adoption. With only about 300,000 AI specialists worldwide, many organizations struggle to acquire the necessary expertise to develop, implement, and maintain AI systems.

To address this skill deficit, organizations need to invest in training and upskilling programs to develop in-house AI expertise. Collaborating with educational institutions and industry partners to create AI talent pipelines is crucial. Also, consider partnering with AI consultants and vendors who can provide the necessary support and guidance.?

By proactively building a strong AI talent base, you can ensure that your organization has the skills and knowledge needed to successfully implement and scale AI initiatives.

5. DATA QUALITY AND INFRASTRUCTURE CHALLENGES

Effective AI implementation relies heavily on high-quality data and robust infrastructure. However, many companies deal with inadequate data management systems, limited processing power, and a lack of suitable tools to handle the large, structured datasets required for AI.

A study by Deloitte found that at least 4 out of 10 companies have low-quality data, and almost a third of the executives said it's one of their biggest challenges with AI.

Another major data quality challenge is the presence of data silos. When data is scattered across different departments, systems, and formats, it becomes difficult to integrate and utilize effectively for AI implementation. Inconsistent data formats, missing values, and duplicate entries can also compromise the accuracy and reliability of AI models.

On the infrastructure front, legacy systems and outdated hardware may not have the necessary processing power and storage capacity to support data-intensive AI workloads. This can lead to slow performance, limited scalability, and increased costs associated with maintaining and upgrading infrastructure.

To overcome these hurdles, start by conducting a thorough data and infrastructure audit to assess your current setup. This "health check" will help you identify areas that need improvement.

Once you've identified the areas that need improvement, invest in modern data management systems and infrastructure that can keep up with the demands of AI. However, it's not just about the technology; you must also establish rules and processes to ensure that your data remains accurate, consistent, and secure.

6. LACK OF AI-FOCUSED VISION

Many organizations have a basic vision for growth: optimizing existing offerings or expanding current products and services. However, to truly harness AI's transformative power, it's essential to develop a transformation vision that aligns with the changing world and anticipates evolving customer needs.

Successful AI transformation requires a bold, forward-thinking mindset. Envision how your organization can leverage AI to create entirely new products, services, and experiences that cater to your customers' evolving needs. Challenge traditional assumptions about your target market and explore the possibility of serving new customer segments that may emerge as a result of AI-driven disruption.

Think beyond incremental improvements and aim for a fundamental reimagining of your business model. Most importantly, recognize that true transformation extends beyond the boundaries of your own organization.

JUMPSTART YOUR AI TRANSFORMATION TODAY

Many companies struggle to overcome the six barriers mentioned above, but there is a proven, data-driven approach to help you get started with AI and achieve your desired business outcomes.?

Our team at FROM: Vision to Victory developed AI Jumpstart, which empowers organizations to take the crucial first steps to AI adoption. We start by engaging key stakeholders in your company to understand your current needs, business goals, and level of experience with AI.?

After assessing your AI readiness, we will then work with you to develop a tailored AI roadmap and establish robust AI governance policies. We can also help you create a robust training plan to prepare your teams for AI implementation.

Contact us today to learn more about AI Jumpstart.

YOUR TURN

What risk or challenge would you be most concerned about when leading an AI-driven transformation in your company? Share your thoughts in the comments below.

SUBSCRIBE to DX Weekly for more content each week to help you and your company succeed in the digital world.

Howard Tiersky is the founder of FROM, The Digital Transformation Agency where he works with leading brands on digital transformation.

• Check out his Wall Street Journal bestselling book, Winning Digital Customers: The Antidote to Irrelevance.
• Howard did a podcast episode with he Founder, Chairman, and Principal Analyst of Constellation Research–a Silicon Valley-based company focused on looking at the future–Ray Wang. Ray talks about the metaverse, decentralized autonomous organizations, Web 3.0 infrastructure, cryptocurrency, and how it will affect all areas of our lives in the coming years. Listen to the episode here.
• And of course, SUBSCRIBE to this newsletter to make sure you stay on top of his weekly insights on digital transformation here on LinkedIn.

Want to learn more? Explore our extensive collection of over a hundred videos on a wide range of topics tailored for executives and directors, from management to digital transformation. Check it out now!