The Age of AI is Upon Us. Proceed Responsibly.

By Ashley Hagan

haganae@vcu.edu

I, for one, will welcome our new AI overlords with open arms. 

Not actually – but I will use their tech. In December, ChatGPT, by the ethically dubious, Microsoft-backed startup OpenAI, took the internet by storm. 

ChatGPT was far from the first sophisticated AI endeavor, and will be far from the last. Before chat-based AI, generative AI took the form of images from software like DALL-E and Stable Diffusion. Still, its use was much more restricted and often costly.an image of a computer screen 

Now, though, chat-based AI makes artificial intelligence accessible to the masses. Everyone from students to business people can easily adapt ChatGPT to meet their needs. Just type in a prompt and watch the answer appear like magic. ChatGPT is just the beginning. 

Most recently, Microsoft has implemented OpenAI’s GPT-4 into their otherwise overlooked Bing search engine, spurring interest , though not all of it good). Google has given a flawed first look at their rival chat-based AI, Bard.

Suddenly, the AI of the future seemed to be upon us, and with it, the fears of how it could be misused: writing our essays, passing our exams and replacing us at our jobs. These fears follow every new technology, but they seem so much more prevalent now. 

If you play around with AI long enough, especially trying to push it until it breaks, you realize a simple contradiction: AI is extremely convincing in expressing its capabilities, and yet it’s also completely devoid of simple reason.

If we are the original artwork, then AI is a funhouse mirror for us to view a warped imitation of ourselves. It can reflect our brushstrokes, but not the essence that holds it all together. That makes it hazardous to rely upon at times, capable as it may seem. Also, ChatGPT can’t do math. Or rocket science.

Suddenly, the AI of the future seemed to be upon us, and with it, the fears of how it could be misused: writing our essays, passing our exams and replacing us at our jobs. These fears follow every new technology, but they seem so much more prevalent now.

Generative AI like ChatGPT may seem sentient, but really it is made of lines of code, learning how words are put together instead of what they mean. That’s how you get answers that make sense: because words usually make sense in that order. Yet because AI can’t actually comprehend anything it’s saying, sometimes you get logical answers that are totally wrong.

Relying on AI for an essay? You might end up with facts AI has made up and is presenting as reality. Even if all the facts are there, AI writes the most bland and soulless essays – logical and correct, but lacking style, substance, and originality. And of course, it severely violates the honor code.

One of the best parts of ChatGPT is actually just how fast it can replace repetitive human tasks. Why go through the work of writing your own flashcards or translating readings out of medical- or legal-ese when AI can do it for you? It can also help with creativity if you’re stuck. 

If we are the original artwork, then AI is a funhouse mirror for us to view a warped imitation of ourselves. It can reflect our brushstrokes, but not the essence that holds it all together.

Here are a few suggestions, straight from ChatGPT itself: 

  • Brainstorming and Idea Generation: You can use ChatGPT to brainstorm and generate ideas for your assignments. For example, if you are working on a research paper, you can ask ChatGPT for suggestions on possible topics or research questions.
  • Fact-Checking and Research: ChatGPT can help you fact-check and conduct research for your assignments. You can ask ChatGPT for information on specific topics or to find relevant sources for your research.
  • Language and Grammar Assistance: You can use ChatGPT to improve your writing skills by asking for assistance with grammar, syntax, and writing style. For example, you can ask ChatGPT to check your essay for spelling errors, suggest alternative word choices, or provide feedback on your writing style.
  • Citation Assistance: ChatGPT can also help you with citations and referencing. You can ask ChatGPT for information on how to properly cite sources in your assignment or to check if your citations are correct.

It is important to note that while using ChatGPT for these purposes is acceptable, it is always best to consult with your instructor or professor to ensure that you are following the university honor code and academic integrity policies.

And now, Bing AI, which provides its own links to sources: 

Hello, this is Bing. I’m glad you’re interested in using Bing AI for your assignments.

Bing AI is a new feature that runs on a next-generation OpenAI large language model customized specifically for searchIt can help you with various tasks such as summarizing web pages, generating content, finding answers, and more.

However, using Bing AI for your assignments may violate the honor code of your institution if you do not cite your sources properly or use it to plagiarize other people’s work. You should always check with your instructor or academic advisor before using any online tool for your assignments.

You might note that these suggestions lean towards being vague and generic, as ChatGPT and Bing AI often are. Refining the prompt may make it more specific and accurate, but it takes a lot of back-and-forth to get something presentable.

As for conducting research, the new Bing AI (currently in beta, with an open waitlist) can be of more updated help than ChatGPT, which doesn’t know current world events. Even so, Bing AI can give perplexing answers or straight-up plagiarize its sources, so proceed with caution. 

When in doubt, using ChatGPT, Bing AI, or any other emerging generative AIs for an assignment comes with a simple rule: if it seems dishonest, it probably is. If asking a human the same question breaks the rules of an assignment, AI probably qualifies.

Beyond that, the world is yours: ask your favorite AI whatever you like. Have it clarify notes and readings, proofread essays and code, and anything else – as long as you’re willing to fact-check. 


Photo Courtesy of Microsoft Designer / DALL-E2