Credit: BeeBright / iStock
When generative AI burst into the public consciousness, many people began to consider it synonymous with ChatGPT, a tool developed by OpenAI with significant investment from Microsoft. But generative AI is also being applied to standard search engines — a market where Microsoft’s main rival, Google, is ubiquitous.
Microsoft’s and Google’s approaches to generative AI are a matter of developing and executing technology strategy. While the technology may be relatively new, the strategy playbook is not, and all companies need to explore their own competitive strategies in the age of artificial intelligence.
In a recent webinar, MIT Sloan professors and who teach the MIT Sloan Executive Education course “Developing and Managing a Successful Technology Strategy,” discussed three ways companies can take advantage of technology development:
Value creation. It often takes years of work in the lab, not to mention several failed experiments, before innovation happens, Azoulay said. For generative AI, that moment happened in late 2017, when Google published the model for transformer architecture. This deep learning model is well suited for natural language processing, and today it’s at the heart of leading generative AI models, Azoulay said.
Converging on a dominant technological paradigm is what allows companies to “make a ton of progress without necessarily investing a lot of money,” Azoulay said, but this takeoff phase must ultimately end before the dominant design becomes a straitjacket for further innovation. Another challenge companies face is bridging the chasm that separates the lead users, who will tolerate bugs for the privilege of being able to say “I was there,” from mainstream users, who need the technology to “just work” and be well supported.
Value capture. Two factors determine whether a firm profits from products it introduces to the market, Azoulay said. One is appropriability, which refers to maintaining control of the knowledge an innovation generates. This may prove difficult: When Meta’s generative AI model, known as Llama, was leaked online, it wasn’t long before open-source models of that technology emerged and produced results that rivaled those of ChatGPT and Google’s Bard.
The other factor is complementary assets, which refers to what a company develops to exploit the knowledge that its innovation generates. Companies will likely focus on two types of assets for creating more value out of their generative AI products, Azoulay said. One is data, because a large language model trained on a unique, proprietary data set may be more valuable than one trained on widely available public data sets. Another is proprietary technology that improves accuracy, reduces bias, and makes models stand out in the crowd.
Value delivery. Whether a company can deliver value based on innovation is less about the technology and more about how the organization responds to it. The challenge is particularly acute in incumbent organizations whose business models might be negatively impacted by the new technology, Zuckerman Sivan said.
Ideally, established companies will not only respond to but anticipate the discontinuity caused by innovation and seek to mine their complementary assets for untapped potential.
Related Articles
But unlike startups, which enjoy a high degree of market responsiveness and autonomy, innovators in incumbent organizations often must check in with other departments and wrestle with tricky questions about competitive outcomes.
“Can you do the old business and the new business at the same time?” Zuckerman Sivan asked. “If you leverage your complementary assets into the new space, do they extend that old business or do they threaten it?”
The good news: Organizations that manage this kind of ambidexterity — doing the old and the new at the same time — often see big gains, including a potential boost to corporate culture. “If I am in a legacy business that’s being helped by new innovation, I’m much more amenable to helping out my colleagues who are in that new business. It’s more of a win-win situation,” Zuckerman Sivan said. “This kind of dynamic is more viable for a company like Microsoft — where AI seems to be more opportunity than threat.”
