Developments in artificial intelligence have broken new ground. Photo courtesy of www.intelligencesquaredus.org
MASON KUPIAINEN | STAFF REPORTER | mkupiainen@butler.edu
As the world continues to develop new advanced technology, society treads murky waters that can create unpredictable dangers. There have been countless recent developments in artificial intelligence that border the line between fascinating and terrifying.
Directed by AI
A prominent development in the world of AI that took social media by storm has been AI-generated art. One of the major companies behind the developments has been Stability.ai, which allows users to get generated portraits of themselves depicted in a variety of situations.
With AI on the trajectory that it is on, it may not seem like there should be cause for concern. However, certain careers may face a significant decline, such as animation, as AI offers cheaper and faster options.
Entertainment giant Netflix faced criticism for an anime short film they released titled “The Dog & The Boy.” The artwork for the three-minute short was generated by AI, with Netflix Japan claiming the film was an “experiment.” Of course, there is potential for more projects with AI contributions, but is there truly a market for AI-created entertainment?
Anna Vadella, sophomore computer science and anthropology double major, shared her stance on artificially made art.
“I think that a big part of my enjoyment from [art] comes from the fact that I know another person put their time and effort into creating this piece of work,” Vadella said. “If we take out that human-to-human aspect of it, then it just feels as the name implies, very artificial. There’s nothing [of] substance to it.”
Colin Harts, a senior majoring in middle secondary education, echoed a similar sentiment.
“I don’t think I’d be interested in [it] the same way as human-produced [art],” Harts said. “I think I’d be interested to see what it looks like.”
The misuse of AI
Amongst all of the exciting potentials AI could make with entertainment, there comes a disturbing aspect. Controversy has abrupted recently after deep fake pornography, digitally created videos or images that make someone appear as someone else, was made of female Twitch streamers without their consent.
“There’s a lot of concern with voices and photos,” Vadella said. “There were [people using AI to take] the face of female streamers and creating AI [deep fake] porn [of them]. It’s things like that where people are now having their identity and privacy taken away in ways that we don’t have regulations for yet.”
Not only can AI create deep fake images of people, but their voices can be synthetically created as well. The company ElevenLabs has created what they call “the most realistic and versatile AI speech software, ever.” This program allows users to generate any voice to say whatever they want.
Advancements in technology that can recreate a human voice bring unimaginable dangers. There comes a real risk of people being accused of making statements they haven’t made.
“We [can] make an AI-generated phrase or sentence [of someone] and say, ‘Oh, you said that because it’s your voice,’” Vadella said. “There’s no way that we can check that right now,”
If this technology continues to improve, it can become difficult to prevent people from taking advantage of it. For example, if someone in court were to create artificial voices of another person to use against them, would there be clear-cut ways of refuting these allegations? If people are adding deep fake images to artificially created voices, could we ever truly trust what we see online?
Written by AI
Entertainment and media have not been the only areas to see an increase in artificial developments. There has been a notable development in AI that could impact education in a crucial way.
EssayGenius has created a way for artificial intelligence to write essays. The website’s blurb states, “EssayGenius uses cutting-edge AI to help you write your essays like never before. Generate ideas, rephrase sentences and have your essay structure built for you.” The website provides a variety of examples of previously written essays on numerous topics including mathematics, philosophy, psychology and many others. These advancements may be neat experiments, but students can take advantage of them in classrooms.
If AI is able to write a well-written essay in a few minutes, this will likely force educators to come up with solutions to prevent students from doing so. Having a concentration in English education, Harts is able to assess the problems classrooms will face.
“It’s a bit of a fearful view because students may be able to use this technology to inevitably write their essays and to turn those in as not their own work, which is obviously not what we’re trying to assess as educators,” Harts said. “Right now, it’s in a place where I’m pretty sure with the reading I’ve been doing about it, that we can detect it and understand whether it’s written by a bot or not. But again, the fear is that students will see it as an easy way out instead of maybe a helpful resource, like finding sources for a paper or something. “
Currently, there are ways for educators to combat the issue. Harts explained how some teachers are falling back on old-school ways of learning, such as having timed, in-class written essays.
Dr. Ankur Gupta, professor of computer science and software engineering, proposes that there are ways to detect if students are turning in artificially written code. This would involve creating programs where instructors can send the assignments through and detect if they were artificially created.
“When writing code, humans have a style that is detectable,” Gupta said. “It’s not that hard to determine if a student wrote the code themselves. [We’ll] just have to come up with solutions to handle it. But I don’t think it’ll be that problematic.”
Developments have been put in place to help handle these rising issues. To help combat the problem, a Princeton University student developed a program that can detect if a paper was written by AI called GPTZero. The program is able to detect human-written essays if the text is complex and has varied sentence lengths, whereas AI-generated essays will include prose that is uniform and familiar.
When science fiction becomes fact
Seeing how artificial intelligence has reached the point where it can write essays and create art can be alarming. Science fiction films like “The Terminator,” “2001: A Space Odyssey” and “Ex Machina” have warned audiences of the potential dangers that come with increasing artificial intelligence. However, these extreme theories of robots taking over mankind may not be as realistic as movies may portray them to be.
Dr. Gupta doesn’t believe there should be any worry with recent AI developments.
“AI is a tool, and it does some things very well, [but it does] only the things we tell it,” Gupta said.
Artificial intelligence, if used wisely, can be a beneficial tool for humanity. It can make lives easier and open up a new world of opportunities. AI can be utilized in healthcare by limiting the spread of disease with contact tracing or analyzing and processing data for research more quickly. Artifical intelligence can also help with mundane tasks, such as sorting through spam emails or operating smart home devices, like Alexa. However, regulations will need to be put in place on these new developments to prevent this technology from being abused.
AI itself isn’t anything to fear, but rather humanity’s potential treatment of the tool is.
