Skip to main content

Verified by Psychology Today

Artificial Intelligence

How Artificial Intelligence Is Changing Education

Are my students using chatbots on assignments? Could I even tell?

Key points

  • AI is changing both students’ and teachers’ options and behaviors.
  • AI could be used to expand needed tutoring and mental health services for students.
  • AI will not replace human interactions, but may perform routine tasks to free up time.
Infinity Shutter/Pexels
Infinity Shutter/Pexels

    AI stands for “artificial intelligence”—computer programs that emulate human-like thought and language. Programmers code and "train” computers to perform tasks that normally require human intelligence, such as visual recognition, artificial speech, and solving problems.

    You’ve probably already noticed AI at work, when you view online ads based on your prior web browsing. AI algorithms are working behind the scenes to customize the ads and content you see, and a natural language processor (NLP) interacts and responds with you in a conversational manner. You’ve likely interacted with AI chatbots, like online customer service pop-ups. Early models were simplistic (“How can I help you?”), but new AI chatbots deal with complex and nuanced questions, access vast amounts of data online, and use NLP to simulate realistic conversations with customers (“How are you doing today?”). Chatbots are now widely available and used for entertainment, treating insomnia, gathering news or statistics, generating sales leads, product branding, and companionship. (1)

    You may have experienced or heard about chatty AI models, such as ChatGPT (Open AI), or Bard (Google), that produce realistic, human-like conversations. These AI programs interact conversationally, answering questions, and even admitting mistakes, or challenging inappropriate requests. However, in their current forms, conversational AIs still produce unusual and even unsettling responses.(2) Programmers continue to fix problems with their bots—and a recent upgrade, Chat-GPT-4, will likely continue to improve at emulating human-like conversations while admittedly remaining "less capable than humans in many real-world scenarios.” You can try interacting with some conversational chatbots, such as ALICE or others, from open-source sites.

    While there are worries about AI bots collecting data and invading privacy, there is at least one AI that may help protect us: "Take It Down" is a bot developed by Meta to help teens remove unauthorized photos from the internet. This smart AI software scours the internet and locates inappropriate images, and can even generate the demand to take it down. AI is a great tool for automating this task that would otherwise take hours to perform manually.

    Garbage In, Garbage Out

    Like any of us, AI systems make mistakes. The current generation of AI chatbots seem unable to fully simulate “humanness,” including perceiving emotions, or attending to nuanced language cues in human conversations. Many skills that humans take for granted remain difficult for AI programmers. Likewise, AIs might reflect the biases of the programmers. AIs might show stereotypes “learned” from their creators.(3) Programmers and psychologists must continue to work out the bugs if they wish to further humanize AI processing and linguistics.(4)

    As a college professor, I’ve noticed how AI will change how we approach education. These AI chatbots can prepare a report in response to any question. Therefore, I am a little bit concerned. What if my students use an AI bot to write essays and papers for a grade? Are my students using chatbots to do their homework? Can I detect writing by a non-human source? Luckily, a tool by OpenAI (the makers of ChatGPT) claims to distinguish text written by humans from that of a non-human machine. However, the current OpenAI tool is not 100% accurate: in one test, OpenAI only identified 26% of AI-written text as “likely AI-written,” and falsely identified human-written text as AI-generated 9% of the time. Not so good. It is amusing that I am trying to use one AI to detect if students are cheating using another AI. Since AI is not perfect at detecting plagiarism, educators must be more creative in teaching and grading. Perhaps assigning personal reflection questions and hand-written papers?

    Another area in which AI might help students are AI tutoring programs, and AI-assisted mental health services. Virtual tutors could be accessible 24/7, focused on target skills and self-paced—which students like. Students may even prefer an anonymous, nonjudgmental, customized-avatar tutor. AI can perform basic assessments and refer students for in-person interactions with school counselors when needed. AI automated systems, perhaps even disguised as entertaining games, might soon help schools identify students at risk for academic, behavioral, and mental health issues.

    Future posts will continue to explore how AI is changing our behavior. We will examine how AI is changing business, the workplace, transportation, communications, new drug development, and healthcare.

    There are those who’ll say that the AI risks outweigh the benefits; from privacy concerns to the sci-fi notion that the AI robots will take over. That is not realistic. AI is here to stay, and this technology will open up new discoveries in all aspects of life. Any major technology-driven shift will result in jobs lost, and also new careers created. AI will produce positive developments, but this tool must be used carefully to avoid risks to livelihoods and privacy.

    Notice: This post was written entirely by a human author.

    References

    (1). https://www.wordstream.com/blog/ws/chatbots

    (2). https://www.nytimes.com/2023/02/16/technology/bing-chatbot-microsoft-chatgpt.html

    (3). https://www.manhattan-institute.org/press/release-new-report-finds-bias-in-chatgpt

    (4). https://www.digitaltrends.com/computing/the-6-biggest-problems-with-chatgpt-right-now/

    advertisement
    More from Jeffrey N Pickens Ph.D.
    More from Psychology Today