Debate Topic: Will AI bring more creativity and innovation or risks and dangers?

Reflecting on the debate, it occurs to me that there were two different forms of creativity and innovation being referenced interchangeably that day, namely, human creativity and innovation on the one hand, and AI creativity and innovation on the other. The effects of AI on human creativity and innovation were discussed extensively in the first debate on AI’s effect on (tertiary) education. The main point made then regarding the negative effects of AI on education, reiterated in the second debate, was that the delegation of conventionally creative and innovative tasks and processes to artificial intelligence would undermine humans’ capabilities in those areas, while the main point made on the positive effects of AI was that such delegation would allow and encourage humans to come up with new and higher ways of thinking creatively and innovatively.

This is where the distinction between the two forms of creativity and innovation becomes important. Notice that even if AI undermines or dampens human creativity and innovation, it does not mean that AI will necessarily be a negative force for creativity and innovation as a whole. If AI takes over the creative tasks and processes performed by humans, human creativity and innovation will simply be replaced by AI creativity and innovation, and there will be no loss of creativity and innovation. Indeed, what is more likely is that artificial intelligence will not only take over certain tasks which are currently conventionally assigned to human creativity and innovation, but will perform them better and faster.

For example, the modeling and discovery of new molecular structures is an area where AI creativity and innovation may soon overtake that of humans. Another area might be graphic design, where image models such as Midjourney and DALL·E have already demonstrated the ability to quickly and reliably generate industry-ready designs. In these cases, it must be said that artificial intelligence in fact serves to increase the level of creativity and innovation, albeit by taking over and displacing human creativity and innovation.

The more optimistic sides on both debates drop the “albeit”. They hold that artificial intelligence will increase and accelerate human creativity and innovation precisely by replacing humans in current conventional creative tasks, thereby forcing humans to explore new avenues of expressing creativity and new methods of innovation. For example, with the help of models that generate new kinds of protein structures, scientists may come up with new and innovative ways of creating and testing drugs. Or, with the help of image models, artists and designers may conceptualize new ways of producing and evaluating art, or invent new artistic mediums through which human creativity and innovation can be expressed.

Making the distinction between human and AI creativity and innovation cannot immediately tell us whether AI will bring more creativity and innovation than risks and dangers, but it does seem to help the pro side, as it shows that even if humans do not pivot to new ways of creating and innovating upon being displaced by artificial intelligence, the total level of creativity and innovation at least does not decrease.

The other thing I found curious was the seemingly implicitly made assumption that creativity and innovation are more or less disjoint from risks and dangers. As members on the con side––that is, on the side which argues that AI will bring more risks and dangers than creativity and innovation––noted, AI can be and is already being used by militaries to automate and optimize operations, develop more effective weapons, and implement strategies designed for new and modern arenas of warfare. Should bad actors follow suit, if they have not already, this would certainly pose significant risks and dangers. The thing is, the upgrades which artificial intelligence offers to militaries can also be classified as creativity and innovation. Just because the purposes are controversial and unpleasant does not mean that we can ignore or take away from potential ingenuity. Indeed, historically, many technological advancements, such as the creation of the first computer or the internet, were inspired by the necessities of war. The application of AI to warfare is dangerous and risky precisely because AI will bring plenty of creativity and innovation to the war room. The recognition that certain risks and dangers come in the form of creativity and innovation would greatly bolster the pro side’s position, as they would be able to claim a significant amount of the con side’s examples and cases as their own.

This is not to say, of course, that all types of risks and dangers which AI might bring come in the form of creativity and innovation. If bad actors get their hands on artificial intelligence which has the knowledge and expertise to build existing types of weapons of mass destruction, and simply use it to build and use such weapons, then it would not be an expression of either human or AI creativity or innovation, but would simply present a significant risk.

I would like to touch upon what I take to be the source or root cause of both the excitement and fear over the potential of artificial intelligence. It appears to me that the way most scientific disciplines work is that there are a group of practitioners, with certain methods and ways of thought in hand, who then apply those methods and ways of thought towards some object of inquiry, be it the stars, the human body, or subatomic particles. In that way, it is stable, since there is a subject––the practitioners, an object––the objects of study, and a relation or action between the subject and the object––the scientific method. The difference, when it comes to artificial intelligence, and of the cognitive sciences––or mind fields––in general, is that the object of study is the method or way of thought itself. Cognitive scientists apply their methods and ways of thought to their methods and ways of thought. Since they study the process of discovery, any discovery they make changes the process. It is a fundamentally unstable, self-referential, self-driving field. Artificial intelligence goes one step further, and seeks to automate this process of self-discovery. I think it is this special feature of artificial intelligence which so spooks the signatories of the Managing AI Risks in an Era of Rapid Progress, and so excites the signatories of Joint Statement on AI Safety and Openness. The possibility of artificial intelligence spiraling out of control is great, but in another sense, that is the very goal of artificial intelligence—to create a program which, like humans, has the ability to think about its own thinking, and so adjust and improve its thinking accordingly.