The Top Five Mistakes Companies Make With AI
Artificial intelligence (AI) is on a serious roll. Consider these eye-popping numbers:
• According to one Gartner report, AI implementation has grown 270% in the last four years. According to another, leading organizations plan to double their number of AI projects in the next year.
• Spending on AI hardware and software is expected to soar from the current level of $37.5 billion to $97.9 billion in 2023, IDC projects.
• A PwC study determined that AI could drive a 14% increase in global GDP by 2030 — the equivalent of an additional $15.7 trillion — due to productivity gains and demand for new products and services. That makes AI the biggest commercial opportunity in today’s economy.
Every business leader is asking what AI means for their company and how they can capitalize on it to gain a competitive advantage, serve customers better and foster high-performing workforces. Many want to get ahead of the trend and stake out ground as front-runners.
However, organizations need to be careful not to rush into AI without an awareness of the common pitfalls that can slow, limit or even ruin their AI initiatives. I’m starting to see some instances of companies falling into five common traps:
1. AI Washing
This is an annoying phenomenon where a company positions an offering as involving AI when, in fact, that’s a stretch. In many AI washing cases, a company is really using more basic forms of data analysis to make something more “intelligent.” This is not true AI, which can learn and take action in response to the data someone sends it and, over time, become smarter and more autonomous as the amount of data increases.
MMC Ventures, a British VC firm, examined 2,830 AI startups in 13 European Union countries and found that a whopping 40% don’t actually use AI in a manner “material” to their businesses.
Given AI’s momentum, it’s no surprise that some organizations around the world would latch onto the hype even while lacking an authentic AI story. But it’s a dangerous tack that puts the company’s credibility at risk and, on a broader level, wrongly confuses the market about what AI is and isn’t.
2. The Shiny Object Syndrome
This is different from AI washing in that an organization may be enthusiastically developing genuine AI capabilities, but it has put the cart before the horse by failing to delineate what important business issues or customer pain points it’s looking to AI to solve.
As with AI washing, it’s an understandable predicament. AI is white-hot, and everyone wants in. But it’s also no different from any other technology since the beginning of time in that a solution in search of a problem is doomed to fail.
Thus, the very first step for any company diving into AI must be to carefully identify and weigh the specific areas of the business that AI has strong potential to improve. Remember: AI is a means to an end, not an end in itself.
3. Architectural Snag
Even after organizations have successfully homed in on appropriate AI applications, some are surprised to discover that their architecture isn’t well suited to AI. For a product development organization, “Do I have the right architecture” is a critical question to be answered. The outcome of AI is only as good as the quality and quantity of data it ingests. If an organization just runs some AI algorithms on top of a system and they do not provide the right data, the AI engine will yield substandard results.
4. Organizational Boundaries
Silos are excellent for storing corn; they’re a curse for an enterprise AI strategy that inherently depends on data integration and collaboration among teams to ensure end-to-end solutions.
In a McKinsey survey, only 17% of respondents said their companies “have mapped out where, across the organization, all potential AI opportunities lie. And only 18% say their companies have a clear strategy in place for sourcing the data that enable AI work.”
Also, companies must make sure they are organizationally and culturally set up to do AI right. For example, a customer support team that is vacuuming up data in their crucial area must be seamlessly sharing that data and collaborating with the data scientists who are writing AI models aimed at improving customer service. That’s the only way to ensure the continuous feedback loop that is essential to AI.
This article originally appeared on forbes.com To read the full article and see the images, click here.
Nastel Technologies uses machine learning to detect anomalies, behavior and sentiment, accelerate decisions, satisfy customers, innovate continuously. To answer business-centric questions and provide actionable guidance for decision-makers, Nastel’s AutoPilot® for Analytics fuses:
- Advanced predictive anomaly detection, Bayesian Classification and other machine learning algorithms
- Raw information handling and analytics speed
- End-to-end business transaction tracking that spans technologies, tiers, and organizations
- Intuitive, easy-to-use data visualizations and dashboards