In this roundup, we dive into research and insights from the Kellogg faculty about the past, present, and future of artificial intelligence and its implications for scientists, advertisers, and the workforce.
1. An overnight success that took decades
While the ubiquity of artificial intelligence may seem like it happened suddenly and very recently, its foundations were laid in the last century by mathematicians and engineers.
“We’ve accomplished a lot, we’ve come a long way in artificial intelligence,” said economics professor Kellogg Sergio Rebelo noted in a recent webinar on the past and future of artificial intelligence. “But this progress only happened after many years of failure.”
Rebelo described how false assumptions about how to build computer programs with expertise hindered his progress. And he noted that this progress was possible because of continued research funding.
“We are where we are because, despite failure for 50 years, the government has continued to fund this research,” he says.
Despite the development, artificial intelligence still has many issues to be resolved. The illusion that AI is part of the information it provides is common. But from Rebelo’s point of view, deciding to stop using AI tools out of fear is a mistake—one that will set people back further.
“There’s a lot of anxiety about AI replacing humans,” says Rebelo. “I’ll tell you – the first people to be replaced are those who don’t know how to use AI. And they’re going to be replaced not by AI but by people who know how to use AI.”
2. Time to train researchers
While AI is often discussed as a game-changer for business, some believe its most significant impact may be in another setting: the laboratory.
“I would argue that the central question in AI today is whether AI can make new scientific discoveries, which is usually seen as the critical milestone toward AI general intelligence,” he says. Dashun Wangprofessor of management and organizations at Kellogg, where he also directs the Center for Science and Innovation (CSSI) and co-directs it Ryan Institute on Complexity.
To better understand how AI is currently benefiting science and how it will benefit science in the future, Wang and Jian Gao, a former assistant research professor at CSSI, assessed how scientists have used the technology over the past decade.
They found that while artificial intelligence is widely used and influential, its benefits are unevenly distributed — with fewer benefits going to fields that have higher proportions of women and minority researchers, such as sociology.
They also found a huge gap in which industries are ready to take advantage of AI. Outside of three core computing disciplines—computer science, mathematics, and engineering—the disciplines were not investing enough in teaching AI-related skills to graduate students and junior scientists to realize the full benefits of AI.
Their research showed a deep reliance on peers who have specialized knowledge to bridge the AI education gap. This suggests that fully exploiting AI in science could require not only more funding to train scientists, but more opportunities for interdisciplinary collaboration.
“Some institutions are launching interdisciplinary research centers, which encourage faculty from different disciplines to have conversations about how to leverage different AI tools and advances,” Gao says. “This would give researchers more exposure to each other to potentially learn from each other when they’re actually doing research together.”
3. The real and the fake
As simple as it is to create realistic images using AI, it can be more difficult to identify which online images are authentic and which are created by AI. But diffusion models that turn text messages into images tend to create what Matt Groh calls “artifacts and oddities” that can provide viewers with clues about where the images came from.
“These models learn to denoise images and generate pixel patterns in images that match text descriptions,” he says. Grohassistant professor of management and organizations at Kellogg. “But these models are never trained to learn concepts like spelling or the laws of physics or human anatomy or simple functional interactions between the buttons on your shirt and the straps on your backpack.”
Grohalong with colleagues Negar Kamali, Karyn Nakamura, Angelos Chatzimparmpas, and Jessica Hullman, recently used three text-to-image AI systems (Midjourney, Stable Diffusion, and Firefly) to create a curated series of images illustrating some of the the most common artifacts and rare.
So what should you look for in an image to determine if it might be artificially created?
As you study an image that you think may have been artificially created, taking a quick inventory of a subject’s body parts is an easy first step. AI models often create bodies that can look unusual — even fantastical. So if the people in your image have extra limbs or digits, a face with hollow eyes or oddly shaped pupils, overlapping teeth, or other anatomical inaccuracies, you may be looking at a fake.
Other signs include unrealistic skin coloring or textures, over-perfected portraits, mismatched lighting, unusual text, glitches in the appearance of common objects like backpacks and ladders, and implausible interactions between people and things. Noticing these unusual details gives you a chance to slow down and check for further evidence that the image you’re seeing could have been created by AI.
“We as humans operate in context,” says Groh. “We’re thinking about different cultures and what would be appropriate and what’s a little bit like, ‘If this actually happened, I probably would have heard about it.’
For more tips and examples, Groh and his team put together a handy reference which is worth studying.
4. How artificial intelligence can shape the future of work
Despite the push to incorporate AI into a variety of workplace settings, it may be a long time before AI comes to your or anyone else’s job.
“Decades of research show the fear is unfounded,” says Kellogg’s Hatim Rahmanassistant professor of Management and Organizations.
In a recent The Insightful Leader Live webinar, Rahman described how integrating AI into our work lives may not happen overnight.
“It will take a long time to penetrate an industry, especially in ways that will affect your career,” says Rahman. A lot will depend, therefore, on how we choose to develop it. And there is time to make that choice collectively and intentionally.
We can choose to use AI to replace as many workers as possible—or we can instead choose to use AI to enhance talent and identify it in unrecognized places. We can choose to let machines make most of the decisions about our health care, education, and defense—or we can choose to keep humans at the helm, ensuring that human values and priorities rule the day.
Critical to prioritizing people in this process is giving people across the workforce an opportunity to shape the employment decisions that affect them, “because without diverse voices and stakeholders, the design and implementation of artificial intelligence has [reflected]and it will reflect, a very narrow interest group of people,” says Rahman.
You can watch Rahman’s webinar here.
5. The future of personalized ads
Imagine being moved by every ad you see online as if the advertisers know you, your personality and the things that matter to you.
In the era of genetic AI, marketers will be able to mine our expanding digital footprints to create highly personalized ads, he says Jacob Tinyassistant professor of marketing at the Kellogg School of Management.
“This particular influence tactic, personalized persuasion, will now be more scalable and widely deployed than ever before,” says Teeny.
Teeny and his researchers recently designed a series of studies to see how customers reacted to personalized presentations written by ChatGPT. They found that the AI was able to write messages tailored to people’s complex psychological profiles, and these messages were more persuasive than non-tailored or generic messages.
Even when participants in one study were informed that the ads they were seeing were generated by artificial intelligence, this did not change the effectiveness of the messages. This is notable because previous research has shown that people try to resist being influenced when they know someone is trying to persuade them.
“We’ve seen a cultural shift in terms of accepting being persuaded. When we go online, we know we’re getting targeted advertising. And because it’s become so commonplace, I think people are more okay with it,” Teeny said.
While marketers may see this as an opportunity to lean into deep personalization, for consumers, this new messaging threshold requires them to be even more careful about the content they read online.
“We’re going to be inundated with things that naturally feel appealing to us. So we may need to take a second step to really investigate the source or accuracy of the message—whether it’s a consumer product or a political news article,” says Teeny.
