3 Major Mistakes Companies Are Making With AI That Is Limiting Their ROI With so many competing narratives around the future of AI, it's no wonder companies are misaligned on the best approach for integrating it into their organizations.
By Chris Stegner Edited by Maria Bailey
Key Takeaways
- 1. Underestimating and limiting AI's potential
- 2. Trying to mimic another company's AI use case
- 3. Buying off-the-shelf products — not tailoring AI solutions to your needs
Opinions expressed by Entrepreneur contributors are their own.
I was talking to a friend recently who serves as the CTO at a mid-sized company and was struck by his sudden change in perspective on AI. Despite initial skepticism, he now believes artificial intelligence (AI) will revolutionize his industry. Yet, his main challenge has been convincing the rest of his executive team to adopt an AI roadmap. This scenario isn't isolated.
In the last year, we've seen a contracted hype cycle around AI, which has caused many leaders to question if an investment in AI can truly yield proportional returns. These concerns aren't without merit. VC firm Sequoia Capital recently estimated the AI industry spent $50 billion on Nvidia chips to train AI models last year, yet only yielded $3 billion in revenue.
Despite that disparity in investment, Sequoia went on to hypothesize AI is likely "the single greatest value creation opportunity" mankind has ever known, comparing its impact on business to that of the cloud transition. Unlike the cloud, however, which replaced software, AI has the potential to replace services, which the VC firm estimated has a total addressable market in the trillions. It's the reason tech giants like Microsoft and Amazon continue to double down on AI investment.
Related: What Is Artificial Intelligence (AI)? Here Are Its Benefits, Uses and More
With so many competing narratives around the future of AI, it's no wonder companies are misaligned on the best approach for integrating it into their organizations. The problem is most leaders are still looking at AI in its limited capacity as a software or tool rather than its ability to operate in a human-like capacity. Here are three common mistakes I see companies make when it comes to implementing an AI roadmap.
Underestimating and limiting AI's potential
AI is widely viewed as a tool or software, but because it can create and reason, it has the ability to interact in a human-like capacity. Much like a junior employee who gets better at their job with experience, AI has the ability to learn from its interactions and refine its methods to improve its output and take on more work overtime.
For this reason, leaders who think of leveraging AI as "smart people" rather than software are better positioned to harness its full potential. Think about a company's organization chart. If you were to write down the skills and tasks associated with each employee, then you can start to visualize where AI can be trained to augment or automate these tasks.
AI already outperforms humans in areas such as image classification, visual reasoning, and even English understanding, according to Stanford University's recently published AI Index report. As of 2023, the report showed AI has surpassed human-level performance on several benchmark tasks, succeeding in helping workers become more productive and produce better-quality work. Another study out of the University of Arkansas showed AI outperformed humans in standardized tests of creative potential.
Unlike humans, however, AI scales up effortlessly as business demands increase, handling workloads without the physical and mental limits of humans. Adopting AI in this way means rethinking our team structures and workflows. It involves training teams to work alongside AI to enhance their roles and drive innovation.
This perspective shift is crucial because it allows leaders, who may not be accustomed to deploying technology themselves, to innately understand how to best leverage AI across their entire organization.
2. Trying to mimic another company's AI use case
The more you start thinking of AI as smart people, the more you realize how individual every organization's approach to building an AI roadmap should be. I like to think of AI implementation as the onboarding of new team members who need to fit within the specific dynamics of your company.
Take human resources for example — one company might have 10 people there; another only three, even if they're the same size. This difference isn't just about company size or revenue. It's about how these companies have evolved.
Each business has its own unique structure, culture and needs. In order to realize generative AI's full potential, PwC reported, businesses must take advantage of its capacity to be customized to a company's specific needs and avoid the use-case trap.
Of course, general use cases for AI exist, particularly when it comes to enhancing customer service or sales. But, when you're looking at a deeper integration of AI into a company's operations, the approach needs to be custom-built, not copied and pasted from outside case studies.
Related: I Tested AI Tools So You Don't Have To. Here's What Worked — and What Didn't.
3. Buying off-the-shelf products — not tailoring AI solutions to your needs
There are some great off-the-shelf AI products like ChatGPT, Dalle, and translation tools that solve specific problems within a company. The challenge with investing in a boxed solution for AI is that many leaders fail to see how AI can enhance operations at a systemic level.
The true power of AI lies in its ability to fundamentally transform your operations, not just perform isolated tasks. PwC's 2024 AI predictions report states that many companies will find attractive ROI from generative AI. Still, few will succeed in achieving transformative value from it — the biggest barrier being the inability of leaders to think beyond boxed solutions and reimagine the way they work with AI.
When building an AI roadmap, leaders must first conduct a thorough assessment of their company's processes. This means identifying areas with redundancies, recognizing outsourced tasks that could be automated, and pinpointing where the company invests heavily in human capital. By understanding these dynamics, leaders can tailor AI solutions to their company's needs and transform how they work.
