In this golden age of innovation, artificial intelligence (AI) currently finds itself at the epicenter of disruption, controversy and limitless possibility.
But let’s not forget, AI wasn’t born yesterday. While an alphanumeric soup of dazzling new AI tools and technologies has suddenly swooped into our lives, AI has figured prominently in a growing number of the world’s businesses for some time now.
Back in 2019, an EY study found that 84% of US business leaders considered AI essential to their companies’ success. And in 2020, IBM reported that more than 85% of adopters were reducing operating costs with AI, and companies attributed an average of 6.3 percentage points of direct revenue gains to AI.
Today, AI’s impact is accelerating across more enterprises than ever, driving down costs, boosting efficiencies, and building stronger businesses. And one strand of AI in particular is on our minds. Generative AI represents the branch of AI’s family tree capable of analyzing data that exists in the digital world to create new content—including images, art, audio, and video.
Trailblazing companies like OpenAI and other startups have suddenly made generative AI accessible to, well, everyone on the planet with an internet connection.
But questions and existential debates continue to swirl around generative AI. Will it really become as smart as humans? Is it destined to redefine our role in the workplace? Where in the world and beyond is it taking us?
There’s no question generative AI is in its nascent stage. But I believe there are plenty of reasons to be inspired and encouraged by the sheer magnitude of its potential across the enterprise:
IT’S MOVING US CLOSER TO ARTIFICIAL GENERAL INTELLIGENCE
I know, “generative AI” and “artificial general intelligence” may sound the same, but they’re very different concepts:
Generative AI is the umbrella category for AI that represents the algorithms used to create and resemble human-like original content.
Artificial general intelligence—or AGI—represents the holy grail for AI: the ability for an intelligent agent to match or outperform the entire intellectual breadth and depth of the human mind. It’s the moonshot for AI innovators.
Generative AI’s tools and platforms are presently not at AGI-level, but their stunning ability to look for data patterns has moved the AI ecosystem’s outputs significantly closer to how a human brain works.
IT’S FUELING A GROWING ECOSYSTEM OF MACHINE LEARNING APPLICATIONS
If applications are compelling indicators of an emerging technology’s success, then early generative AI use cases suggest a brilliant—and potentially disruptive—future.
Microsoft recently introduced a version of its Bing search engine that incorporates ChatGPT technology, and it quickly generated plenty of questions. OpenAI’s series of GPT models has been used to simplify all sorts of enterprise-level processes. They’ve been deployed by enterprise search platforms to help companies answer complex questions, to speed the process of analyzing and summarizing user feedback, and to write computer code.
Finally, other potentially transformative use cases are understandably sending shivers throughout the creative world. Apps are promising effective marketing content in mere seconds, while developers are already proving the ways in which GPT-3’s writing skills could contribute to entertainment and education. Imagine a teaching app that lets students “interview” Albert Einstein in a science lesson. Another OpenAI product, DALL-E, creates variations of images based on natural language text prompts.
IT’S DRIVING HEALTHY DEBATE AROUND AI’S FUTURE.
Generative AI’s sophistication and flaws are fueling debate about its future—from the commercial and societal risks of inaccurate, inconsistent, and biased outputs to the more existential risks of machine-generated outputs that pass the Turing test (when a person cannot distinguish between machine and human output).
OpenAI, for example, has worked to address these sorts of concerns through content filters and new versions of its GPT model, but the debates are far from over. Last December, its chatbot application—ChatGPT—reached one million users within a week of its release. But a flood of AI-generated answers hit the coding Q&A website Stack Overflow, instigating a temporary ban because too many were inaccurate but sounded correct.
There’s no question that major concerns will persist around generative AI’s trust, safety, and accuracy. It can’t yet be assumed that what it generates is 100% true or factual, making the technology easy prey for critics. But I have full confidence that the companies orbiting the generative AI space will continue to explore solutions for addressing these and other challenges. For example, OpenAI’s GPT-4 model has shown a 40% increase in providing accurate response compared to its previous GPT-3.5 model.
I also support conversations around the capabilities of these tools while the technologies are still in their adoption cycle. Asking the hard questions early on based on their real-world applications will be invaluable. They’ll help us seize generative AI’s incredible potential while safeguarding against risks and promoting transparency.
EXPLORING INFINITY AND BEYOND, RESPONSIBLY
When discussing generative AI’s journey—and AI’s broader ecosystem—perhaps what’s most remarkable is that it’s just beginning. That’s why it’s so crucial that users continuously examine and pressure-test the impact of AI tools across the enterprise.
It’s important to experiment early and often: to pinpoint where AI-powered technologies can thrive inside their enterprise and where they could fail. Any learnings gathered now will help businesses accelerate into the AI-powered age. However, as with all emerging tech, you should access and experiment with caution, placing a heavy priority on information security and data privacy, for example.
And finally, as industries explore new ways to leverage these technologies, you must place humans at the center of your efforts to ensure AI works hard and with great purpose for your customers and society.
Yes, these innovations are boldly reaching for infinity and beyond, at light speed. But you need to stop and celebrate the fact that human ingenuity has created such a stunning spectrum of technologies that can so closely imitate human language capabilities. It’s a remarkable achievement.
These are remarkable times.
Jeff Wong is the Global Chief Innovation Officer of Ernst & Young, one of the largest professional services organizations in the world.