Breaking the generative AI hype cycle

Even by the relatively high standards of the technology industry, the hype surrounding generative AI may have broken new ground. McKinsey estimates that this technology could create between $2.6 trillion and $4.4 trillion in annual economic value across industries ranging from banking to life sciences. Chat GPT, perhaps the most well-known example of generative AI, reached 100 million users two months after its launch, and most estimates now have more than 1 billion users. The analysis below is based on several recent studies and the framework for analyzing the impact of automation on jobs described in Reinventing Jobs (HBR Press, 2018), and is based on RPA and currently available shows the potential significant benefits of using four generations of AI. Us (rule-based systems, machine learning, deep learning, generative AI).

But how exactly does generative AI impact work? To answer this question, you need to understand the difference between generative AI and previous generations of AI.Analysis below¹ Shows the most important differences

While early iterations of AI helped predict specific outcomes, the lack of reasoning underpinning generative AI means that its well-documented illusions will remain a feature of this technology for the foreseeable future. will continue to be. To understand how to best use it, you need to understand the essential characteristics of the job. As explained in “Reinventing Work'', every task has four different potential outcomes associated with it.

• Eliminate errors — think of the job of an airline pilot. The impact of mistakes is high, and deviations from acceptable levels of performance are very likely to result in negative value to the organization.
• Minimize variances such as transaction processing work where increasing performance beyond the target level is not worth it.
• In productivity improvements (for example, in salespeople's jobs), improved performance results in commensurate increases in organizational value.
• Achieving breakthroughs — think of highly creative jobs such as data science — where small improvements in performance can have an exponentially larger impact on value.

Established automation, such as robotic process automation (RPA), aims to reduce variance and can help replace human effort where there is a higher tolerance for risk. Consider applying RPA to reduce variance in the highly repetitive, rules-based tasks of analyzing and synthesizing financial data. Previous generations of AI have long been used to power analytical work, where the goal is to increase productivity or achieve breakthroughs. Oncologists will be trained on large amounts of specific data and images to dramatically increase the accuracy of cancer detection, not by replacing their skills, but by enhancing their capabilities and increasing the premium for experience and expertise. Let's consider how we used machine learning. However, if error removal is the objective function of the job, generative AI can be a very risky proposition. The aforementioned illusion places too much emphasis on human ability and consideration, significantly reducing its value proposition.

It is essential for leaders to understand when to rely on various technologies and when not to rely on them, and the specific role technology should play in replacing or enhancing human work to transform it. Generative AI is democratizing knowledge and creativity through augmentation, achieving increased productivity and placing a premium on skills traditionally required for a variety of creative tasks aimed at breakthroughs in areas with high risk tolerance. Most useful for lowering. If you have a high tolerance for risk, it can be a valuable tool for increasing your productivity. Attractive use cases range from copywriting to call center operations, and it has proven to be of great value in improving productivity, especially for less experienced talent.

Interestingly, this democratization of access is similar to the automation of the Second Industrial Revolution, as generative AI improves the productivity of low-skilled talent, and the disadvantages we experienced in the last two industrial revolutions. It may have the benefit of mitigating the expansion of equality. As AI lowers the premium on creativity and democratizes access, how can we continually reinvent our business models and our workforce to leverage it in thoughtful and nuanced ways?

1) Source:

•Generative AI: How it works, history, pros and cons, Investopedia, 2023

•How Generative AI Will Change Creative Work, Harvard Business Review, 2023

• “Generative AI at Work”, NBER, 2023