Yes, Artificial Intelligence Has A Creative Side, Sort Of – Forbes
Not magic, but human creativity accentuated by machines
Despite perceptions, artificial intelligence doesn’t have a shred of creativity within it. It’s all statistical algorithms, slurping data from sources created by humans somewhere along the line. It will never be a source of innovation, but will serve to augment human innovation.
“AI will not develop fundamentally new ideas on its own; however, there are ways in which AI can support humans in doing so,” according to a recent study out of the Gottlieb Duttweiler Institute.
In this context, AI serves as a liberating force. “Foremost among AI’s many capabilities for augmenting human creativity is its ability to relieve us of monotonous tasks, such as searching for information and allowing more time for stimulating activities that can lead to new ideas,” the study’s author, Jan Bieser, points out. “AI can also take over more creative tasks by identifying patterns in data that humans would not have found. AI does not just take over tasks that would be time-consuming; it might provide insights humans would have never found themselves.”
There’s a limit to AI’s ability to assist innovation, however. “AI cannot support all human skills that are essential for idea development, such as real-life observations or personal interactions,” Bieser says. “In addition, exploration without a dedicated outcome in mind, adding new domains of knowledge on the go and improvisation are challenging for AI. For example, a tool that optimizes flight routes in terms of CO2 emissions would not simply suggest that we switch to transportation via trains or meetings via video conference.”
Industry experts agree that innovation cannot — and may never happen — by AI alone. But the potential is there. “Care needs to be taken to fully depend on AI for any tasks,” says Andy Thurai, analyst with Constellation Research. “At this point AI is still not reliable, biased in many ways, not very explainable, and not very accurate in many areas.”
While AI “may not be able to innovate things by itself at the current time, but using AI assistance can definitely assist humans with mundane, repeatable, high-volume tasks and allow humans to think creatively,” Thurai points out. Start with basic data crunching. “Data volumes at current levels are beyond human comprehension. For data quality, data mining and pattern identification with current volumes of data it will take enormous amounts of time for humans to figure it out. AI can help in these areas allowing humans to concentrate in applying those patterns to identify possible solutions.”
There are other compelling examples of AI-enhanced innovation in action. “Drug discovery using various chemical combinations and trials can be sped up with AI,” says Thurai. “As the development of Covid vaccines demonstrated, we don’t have to wait for the conventional decades to develop a vaccine, it can be done much faster using AI.” In another example of sophisticated innovation, human genome sequencing would have been near impossible without AI, Thurai adds. “This allows us to know so much about every individual and find cure that can be individualized rather than a mass treatment using common medication that may or may not work.”
To empower human innovators with AI, Bieser makes the following recommendations: