It turns out Pop stars Drake and The Weeknd didn’t suddenly drop a new song trending on TikTok and YouTube in April 2023. The photo that won an international photo contest that same month was not a real photo. And the image of Pope Francis wearing a Balenciaga jacket in March 2023?
All of this was created with the help of Generative Artificial Intelligence, a new technology that can generate human-like text, voice, and images on demand through programs like ChatGPT, Midjourney, and Bard.
There is certainly something unsettling about how easily people can be fooled by these fakes. I believe this is a precursor to an authenticity crisis that raises some difficult questions.
How will voters decide if a video of a political candidate saying offensive things is real or AI-generated? AI can now create something visually stunning When are people willing to pay artists for their work? Why follow certain authors when stories in their style circulate freely on the Internet? ?
I’m a professor at Stanford University, where I also lead large-scale generative AI and education initiatives.
Text, images, audio, and video are all becoming so easy for anyone to create with new generative AI tools that I think we need to rethink and recalibrate how we determine authenticity in the first place. .
Fortunately, the social sciences offer some guidance.
Long before generative AI and ChatGPT, people were looking for what made something feel real.
When a real estate agent is making a fuss about a property they’re trying to sell you, are they real or just trying to close the deal? That stylish acquaintance wears real designer fashion or wearing mass-produced imitations? How do you discover your true self as you mature?
These are not just philosophical exercises. Neuroscience studies have shown that believing a work of art to be real activates reward centers in the brain, just as seeing something said to be fake is not.
Credibility is also important, as authenticity is the social glue that strengthens trust. Consider the social media disinformation crisis, where fake news is inadvertently spread and real news is declared fake.
In short, authenticity is important both for individuals and for society as a whole.
But what actually makes something feel real?
Psychologist George Newman explored this issue in a series of studies. He found that authenticity has three main aspects.
One of them is historical authenticity, whether the object really belongs to the point in time, place, and person someone claims to be. The actual paintings painted by Rembrandt have historical authenticity. Modern counterfeiting is not.
The second dimension of authenticity comes into play, for example, when a Japanese restaurant offers exceptional, authentic Neapolitan pizza. Their pizza is neither made in Naples nor imported from Italy. The chef who created it may not have a drop of Italian blood in it. But the ingredients, the look, and the taste may be very much in line with what tourists expect from a fine Naples restaurant, Newman calls it categorical authenticity.
And finally, there is authenticity that comes from our values and beliefs. This is the kind of thing many voters find lacking in politicians and elected leaders who say one thing and do another. .
In my own research, I’ve also found that authenticity can be related to our expectations of what tools and activities are needed to make things.
For example, when you see custom furniture that claims to be handcrafted, you probably think it’s not literally made by hand, but uses all sorts of modern tools to cut, mold and attach each piece. Similarly, if an architect were to use computer software to create plans for a building, they would consider the product to be legitimate and original. This is because there is a general understanding that those tools are part of what is needed to make those products.
Most of the time, we don’t think much about these aspects when making a quick judgment of authenticity. But with generative AI, we have to do it.
At a time when it took a lot of time to create original new content, creation required skill and could only be created by skilled individuals putting in a lot of effort and doing their best. Because there was a general assumption that it could. intention.
These are no longer safe assumptions.
Generative AI succeeds by capitalizing on people’s reliance on absolute believability by producing material that looks “real”.
Therefore, it becomes important to disentangle historical and assertive authenticity in one’s thinking. Just because a recording sounds exactly like Drake’s, i.e. fits categorical expectations for Drake’s music, does not mean that Drake actually recorded it. A great essay submitted may not actually come from a student struggling to create sentences on a word processor for hours.
If it looks like a duck, walks like a duck, and quacks like a duck, you should consider that it’s possible that no one actually hatched from an egg.
It’s also important to understand what these new generative AI tools can and can’t really do. This includes ensuring that people learn about AI in schools and workplaces, and having open conversations about how the widespread availability of AI will change the creative process. think.
In the future, writing papers at school does not necessarily mean that students must meticulously craft every sentence. Now we have tools that help us think about how to express our ideas. And you don’t need extraordinary eye-hand coordination or mastery of Adobe Photoshop or Adobe Illustrator to create great photos.
Finally, in a world where AI acts as a tool, society should consider how to establish guardrails. These may take the form of regulations or create norms within specific areas to disclose when and how AI is used.
Are AIs allowed as co-authors of writing? Are they not allowed on certain types of documents or at certain grade levels in schools? Do you want a signed statement that you don’t use AI? Or do you want another new competition that explicitly invites AI-generated work?
These questions are tricky. It may be tempting to simply view generative AI as an unacceptable aid, in the same way calculators are banned in some math classes.
However, isolating new technologies risks imposing arbitrary limits on human creativity. If photography had been seen as an unfair use of technology, would the expressiveness of images be what they are? What if I am considered ineligible for
The capabilities of generative AI surprise many and challenge everyone to think differently. But we believe humans can use AI to push the boundaries of what is possible and create interesting, valuable, and certainly authentic works of art, writing, and design.
Victor R. Lee is Associate Professor of Learning Science and Technology Design in Education. Stanford University
This article is republished from The Conversation under a Creative Commons license. Please read the original article.
