Snapchat Inc. has unveiled My AI, powered by ChatGPT and available for $3.99 per month to Snapchat Plus subscribers - joining the growing field of generative AI technology.
This bot will be located within Snapchat's chat tab and available exclusively to Plus members for $3.99 monthly subscription fee.
My AI will be a speedy mobile-friendly version of OpenAI's ChatGPT that Snap employees have trained to comply with its trust and safety guidelines. This means it cannot answer questions containing swearing, violence, sexually explicit content, or opinions about contentious topics like politics.
Furthermore, Snap notes that My AI only responds to certain queries and does not compose essays.
My AI is limited in its capabilities, but it can still fulfill basic requests. For instance, it could provide suggestions about gifts for friends.
However, it won't have the capacity to tackle complex questions about the world's issues. Instead, it will focus on more personal matters like what to cook for dinner or how best to commemorate a special occasion.
Snapchat's My AI feature will initially only be accessible to those who subscribe to its premium service, but eventually will be accessible to all 750 million monthly users as CEO Evan Spiegel announced.
Convincing AI chatbots to behave is one of the biggest challenges for companies creating them. Alphabet's Google, for example, experienced a major setback earlier this month when its latest chatbot shared inaccurate information in an advertisement video.
Schools are concerned about students using technology to cheat on assignments. Some institutions have banned it from internet networks and school devices, while others allow it for research purposes.
Some educators have noted the software can be beneficial for students who require quick essay production or who write in a foreign language. University of Arizona professors assert that while ChatGPT may enable plagiarism, it also presents an opportunity for learners to push their learning beyond its current boundaries.
Another issue with ChatGPT is its incapability to be detected by plagiarism safeguards. Its output often looks so similar to human students' work that it's difficult for a human assessor to differentiate them. To address this problem, the New York City Department of Education has attempted to ban chatbots from city's internet networks and school devices.
On Friday, Sciences Po, a French university, issued a ban on ChatGPT use in public schools. They issued an email to all staff and students alerting them of their choice.
The school system's ban, which applies to internet networks and devices owned by the system, came after criticizing the program for producing "coherent prose," such as articles, essays and even poetry.
As a result, some teachers are fearing it can be used for cheating on assignments; The New York Times recently showed sample writing produced by the chatbot to writers and educators who could not tell it apart from real work.