AI as Answer Machine? Or Wordsmith? Or Something Else?

Last month on the blog, I floated the idea of thinking about generative AI tools like ChatGPT, Bing, and Bard as technologies for learning. I drew an analogy to the digital camera I used as I learned photography. As I was developing a conceptual understanding of photography, I was also learning how to use the tool in my hands, with all its knobs and buttons, and these two kinds of learning informed and supported each other. Could we use generative AI tools in a similar fashion, teaching students to use the tools well as part of learning other skills, like writing and researching? If so, it might help to know what kind of tools we’re dealing with when we talk about ChatGPT and other AI text generation tools.

For instance, I saw this post by Simon Harris go around Mastodon a couple of weeks ago:

15yo: I use ChatGPT quite a bit.
Me: Oh?
15yo: Today, I asked it to give me the last digit of Pi.
Me: but there is no…
15yo: …exactly
Me:
15yo: Apparently, it’s 4

This is funny because ChatGPT is so, so wrong about pi. As an irrational number, pi has a nonterminating decimal representation; there’s no last digit of pi. The joke, however, relies on an assumption that ChatGPT is supposed to be some kind of answer machine, a tool for answering questions correctly. However, ChatGPT is famously not that. It’s well known for “hallucinating” facts and inventing citations. There are, in fact, a lot of educators who are worried that students will rely too much on ChatGPT and similar tools because these tools produce text that is factually incorrect. For example, check out this YouTube video of a chemistry instructor evaluating the accuracy of ChatGPT’s answers to various organic chemistry questions. Spoiler: It didn’t do that great.

So if these tools aren’t answer machines, what are they? I would argue that ChatGPT and similar tools are actually wordsmiths. They are powered by large language models that mix words together to create fluent prose. Ask it to draft an essay for you on a given topic, and it will put words together to do just that. Ask it to revise that draft based on your feedback, and it will do that, too. See, for instance, how Ethan Mollick used Bing to help him craft the final paragraph of his April 26th newsletter. Ask it to summarize a long text, and, because it’s good with words, you’ll get a pretty decent summary. It’s a wordsmith that actually doesn’t know what it’s writing but still produces intelligible texts.

This is, in fact, the other reason that educators are worried about AI text generators, because they can quickly produce acceptable responses to pretty standard essay question prompts, and educators are worried that students will use these tools to cheat, whether the essay question is of the “writing as assessment,” “writing to learn,” or “learn to write” variety. Because ChatGPT and its peers are good with words.

I’m struck by how the initial reporting and framing of these tools affect the ways we think about them. While AI text generators were seen as answer machines (albeit flawed ones), AI image generators have been seen as “smiths” of a different kind, working with pixels to create images. While a tool like ChatGPT making stuff up is seen as a problem, if I ask Midjourney or DALL-E (a couple of popular image generators) to give me a scene of robots playing basketball, no one worries about the result not being real-world realistic.

If AI text generators are tools for learning, as I argued on the blog last month, then I think it’s helpful to think of them as wordsmithing tools. Does a student struggle to come up with counter arguments for their thesis? Ask an AI tool to generate a few. Does a student freeze up when confronted with a blank page? Ask an AI tool to write a first draft. Does a student find it hard to adopt the right voice for their writing? Ask an AI tool to rewrite a draft essay in different voices. These are all wordsmithing tasks that ChatGPT and its ilk can handle pretty well.

Please note that I’m not suggesting we shouldn’t help students learn to develop counter arguments or to get past writer’s block or to figure out their writing voice. I am suggesting, however, that AI tools might be used to help students learn these skills, just as my digital camera helped me learn photography. Cynthia Alby uses the metaphor of learning to ride a bicycle with training wheels to talk about how AI tools might provide a kind of scaffolding for students learning writing skills. When I heard that metaphor, I was reminded of a 2012 blog post by Mark Sample, in which he talks about balance bikes as a different way to teach kids to ride bicycles. Learning with training wheels means getting help with balancing while learning to pedal, while learning with a balance bike means not worrying about pedaling while learning to balance, which is apparently easier for most kids. We might need to think carefully about where our students need the kind of support that AI tools can provide, and the answer might vary by student.

I’ve been using the training wheels and balance bike metaphors in my recent talks on teaching with AI, and I was all ready to bust out the “AI as wordsmith” argument above at a talk I gave earlier this week. But then I signed up for ChatGPT Pro to see what ChatGPT’s plugins are all about. Back in March, OpenAI, the company behind ChatGPT, announced that the tool would have plugins that allowed it to communicate (in a fashion) with other web tools and websites. It’s taken me three months, but I finally started playing around with these plugins. I quickly realized that some of the advice I had given faculty earlier in the year about designing “ChatGPT-proof” assignments hadn’t been accurate for a while!

For instance, if I ask the free version of ChatGPT, which uses GPT-3.5, to tell me about the May 2023 Supreme Court decision that dealt with Andy Warhol’s art, here’s the response I get:

I apologize for the confusion, but as of my knowledge cutoff in September 2021, I’m not aware of any specific Supreme Court decision in May 2023 related to Andy Warhol’s art. It’s possible that such a decision may have occurred after my knowledge cutoff or it may be a hypothetical scenario that hasn’t actually taken place. I recommend checking reliable news sources or legal references to get the most accurate and up-to-date information on any recent Supreme Court decisions regarding Andy Warhol’s art.

If I think about ChatGPT as an answer machine, I’m not getting the answer I want, but at least the tool is telling me what it doesn’t know. That’s an improvement over how the chatbot presented information at the start of 2023. But watch what happens when I use the same prompt in ChatGPT Pro with the VoxScript plugin enabled:

In May 2023, the U.S. Supreme Court ruled against the Andy Warhol Foundation in a copyright infringement case. The case was centered around a series of silk screen images that Warhol created based on a photograph of the late musician Prince, taken by photographer Lynn Goldsmith in 1981.

The court ruled in a 7-2 vote that Warhol infringed on Goldsmith’s copyright. The case hinged on whether Warhol’s images of Prince transformed Goldsmith’s photograph to a great enough degree to stave off claims of copyright infringement and therefore be considered as fair use under copyright law.

Goldsmith had sued the Andy Warhol Foundation for the Visual Arts (AWF) for copyright infringement after the foundation licensed an image of Warhol’s titled “Orange Prince” based on Goldsmith’s image of the pop artist to Condé Nast in 2016 for use in its publication Vanity Fair…

It goes on for another couple of paragraphs, and it gives me the news source it used to generate this summary of the case, a May 18th NPR report. This version of ChatGPT is indeed acting like an answering machine!

Quick sidebar: It is also now possible to share one’s ChatGPT correspondences with others. A few months ago, you needed some third-party tool to do so, but now ChatGPT has a share button for every chat. Thus the argument that there’s no “paper trail” for how one uses ChatGPT is no longer true.

I used ChatGPT for the above examples, but please remember that ChatGPT isn’t the only game in town. I asked Bing about the Andy Warhol case, and I got a similar summary, this time with three different news sources cited. Google Bard gave me a similar summary and added some text about how fair use works in the United States. Its sources, however, were meeting notes from a 2004 Congressional subcommittee hearing and a Quizlet created for maybe a law course? Your mileage may vary with these tools.

So are these tools now, in fact, answer machines? I’m going to argue that while they can be used in that way, a better way to think of them is as interfaces to other systems. Just as Siri or Alexa act as a voice-based interface to the apps on your digital devices, ChatGPT and Bing and such are text-based interfaces to, for instance, all the transcripts attached to all the YouTube videos. These tools are still wordsmiths, but wordsmiths used as interfaces.

To use a left field example, the fast food chain Wendy’s recently announced that it was working on an AI chatbot to take drive-thru orders. That’s AI-as-interface, with the chatbot, which is good with words, taking spoken orders for a cheeseburger and a Frosty and turning that into an order in the restaurant’s food production system. For a more academic example, consider Elicit, an AI research assistant that can help you find research in its corpus of 115 million academic papers that might address your research question. Ask it a question like “How should I form groups of students when using group work in the college classroom?” and (because it’s good with words) it will return a table of potentially relevant papers, summarized and tagged by intervention, sample size, and outcomes.

AI as answer machine, AI as wordsmith, AI as interface… I’m not sure where we’ll land with how we think about these tools, but I believe these framings are helpful in assessing what the tools can and cannot do and what roles they might play as learning technologies.

One more caveat: Most of what I’ve presented here is focused squarely on AI text generation tools powered by large language models. There are lots of other kinds of AI tools, like Cornell University’s Merlin app, which can identify birds by their songs, or that AI-powered agent that excelled at the strategic negotiation game Diplomacy, that don’t fall into any of the categories I’ve described here.

Chatbot as Audience: A Conversation with Chad

After weeks of reading about other college educators having conversations with ChatGPT, the new artificial intelligence (AI) chatbot from OpenAI that uses the GPT-3 family of large language models, I thought it was time to have a conversation with ChatGPT myself. (Disclaimer: I don’t entirely understand all the terms I used in that last sentence.) See, I’ve been thinking about that thing that Lee Skallerup Bessette said last month about having students interact with chatbots as an audience for their academic work, and I wanted to figure out how that might work. Also, I was preparing to give a talk at Temple University about teaching and technology, and I couldn’t not mention ChatGPT in January 2023, right?

One of the ideas I had about student activities using ChatGPT, which I’ll call Chad from here out for convenience (after Alan Levine), is asking students to talk to Chad about some topic they already know well (a hobby or maybe a sport), to see what Chad does and doesn’t know about things. (I’m seriously anthropomorphizing here, of course. Chad isn’t a person and doesn’t know anything. Chad is an algorithm that I just barely understand.) Chad presents himself as an authority on lots of stuff, but it’s been clear from other educators’ interactions with Chad that Chad gets stuff wrong a lot. That’s a good lesson about AI text generation tools for students, and having them experience that firsthand about a topic they already know might help that lesson sink in.

I occasionally teach a course on cryptography, which includes discussion of the role of encryption in today’s society, which leads us to topics like surveillance and privacy. This a topic I know well, and one that I’ve had many students write papers about, so I chose this topic for my conversation with Chad. Here’s how it started:

Me: There has been much debate over the surveillance powers that governments have over their citizens. But if you have nothing to hide, why would you worry if your government is surveilling you?

Chad: The idea that “if you have nothing to hide, you have nothing to worry about” is a common argument in favor of government surveillance, but it is flawed for several reasons.

First, it presumes that the government’s surveillance is only directed at people who are doing something wrong, but that is not always the case. Surveillance programs can be used to target people based on their race, religion, political beliefs, or other factors that have nothing to do with criminal activity. Even if a person is not doing anything wrong, they may still be subject to harassment or discrimination if they are deemed to be “suspicious” because of the way they look or the group of people they associate with.

Additionally, the idea that you have nothing to hide implies that privacy is only important for those who are doing something wrong. However, privacy is a fundamental human right that is necessary for many aspects of our lives, such as the freedom to express ourselves and to make personal choices without fear of retaliation. It also allows us to establish and maintain relationships with others, and to participate in society without fear of surveillance.

Finally, even if you believe that you are not doing anything wrong and you have nothing to hide, the power to surveil still poses a danger to society as a whole. Governments that have the power to surveil their citizens without checks and balances can easily abuse that power to stifle dissent, silence critics, and punish political opponents. This can have a chilling effect on democracy and the free exchange of ideas, which is essential for a healthy society.

In summary the idea of ‘nothing to hide’ is a red herring, that misses the point that privacy is a fundamental human right, it is not only about hiding something illegal but also about the freedom of choice and self-expression, it’s about the preservation of democracy and free speech as well as for personal safety and security.

Did you actually ready all of that? If so, good for you. Chad is very, very wordy. He talks in solid, if unexciting three-point essays. Which is why so many educators have been freaking out these last couple of months, because if your homework involves assigning students to write solid, if unexciting three-point essays, then you can bet on the fact that Chad has been ghostwriting for your students since November.

Let’s not focus on that, however. Let’s imagine that in the discourse above, I’m a student talking to Chad about a topic I know well, or maybe one that I’m just learning about, like the tradeoffs between privacy and surveillance. I ask a relatively short question, and Chad responds with a longwinded three-point essay. What should I do next, if I’m trying to treat Chad like an audience for my growing understanding of privacy and surveillance? Let me see if I can point out a potential flaw in Chad’s argument. That would be a useful move to make.

Me: Some governments have robust checks-and-balances to prevent the kind of abuse you’re describing. Doesn’t that mitigate the concerns you’ve raised here?

Chad: Checks and balances are an important safeguard against government abuse of surveillance powers, but they are not a guarantee that such abuse will never occur. Even with robust checks and balances in place, there is always a risk that government officials will find ways to circumvent or undermine them.

For example, laws governing surveillance may be written in such a way that they are easily interpreted to allow for wide-ranging surveillance. Or, oversight bodies may lack the resources or independence to effectively monitor government actions. Or, Whistleblowers may be afraid to come forward.

Additionally, even when checks and balances are functioning as intended, they may not be sufficient to protect against certain kinds of abuse. For example, a surveillance program that targets individuals based on their race or political beliefs may not be illegal, but it can still be unjust and harmful.

It’s important to remember that the mere existence of checks and balances does not necessarily mean that a government’s surveillance powers are being used in a way that is consistent with human rights and democratic principles. It’s important to keep monitoring the government’s actions and be vigilant to potential abuse and push for more effective oversight and regulation mechanisms. And also to have transparency, so that citizens can see what their government is doing and have a say in the matter.

It’s also important to have a open public debate on the government’s surveillance powers, the purpose of such powers and the balance between protecting citizens and maintaining individual rights and freedoms. Only then we can have a society where citizens can have trust in their government and vice-versa.

Seriously, Chad, read the room. Short questions need short answers.

When composing my question to Chad, I had a very interesting moment. I typed, “Some governments have robust checks-and-balances to prevent the kind of…” Overreach is the word I wanted to use, but I wasn’t sure if Chad would understand that, so I went with “abuse” instead. Those aren’t entirely synonyms, however, and using “abuse” might take the argument in a direction I don’t want it to go. But would Chad get the nuance that comes with “overreach” in this discussion?

Again, I’m fully anthropomorphizing here, but that’s kind of the point. Chad acts something like an interested (perhaps overly interested) respondent for my questions about this topic. Playing out this exercise is getting me to think about my word choice as I communicate with an audience I don’t know very well. I totally want the students in my writing seminars to think about their word choices like this!

Let’s finish this debate.

Me: It’s true that checks and balances don’t prevent all overreach or abuse, but we citizens already cede a lot of power to our governments, trusting that the benefits of doing so outweigh the costs. You haven’t addressed the benefits of government surveillance for safety and threat prevention.?

Chad: You are correct that government surveillance can play a role in keeping citizens safe and preventing threats. Some argue that surveillance can help to detect and prevent terrorism, for example, or to locate and capture dangerous criminals. In some cases, government surveillance may be necessary to protect the public from immediate harm.

However, it’s important to keep in mind that the benefits of surveillance must be balanced against the costs to individual privacy and civil liberties. The decision to use surveillance should be based on a careful consideration of the potential risks and benefits, and it should be subject to robust oversight and accountability mechanisms to ensure that it is not being used in a way that is arbitrary or unjust.

Additionally, it’s important to consider the potential of alternative methods of threat prevention or safety enhancement which may not involve surveillance or may involve less intrusion to individual privacy. Such alternatives may be more effective and less costly in terms of civil liberties.

It’s also worth noting that some of the information gathered through surveillance may not be as useful as it may seem at first glance. For example, indiscriminate mass surveillance may lead to a large number of false leads and distractions, which can actually make it more difficult for law enforcement to identify and investigate real threats. And the collection of vast amount of data can also raise serious concerns about data security, data breaches and mishandling of information.

Finally, it’s important to remember that surveillance must be carried out in a way that respects human rights and the rule of law, as well as being transparent, proportionate and targeted. This means that the surveillance should be conducted in accordance with the law, and be necessary and proportionate to achieving a legitimate aim.

Okay, now Chad is making some rhetorical moves that go beyond the three-point essay. “You are correct that… However…” Again, not an exciting rhetorical move, but a solid one, and one that some of my first-year writing students struggle to make. I feel like I could do something with this kind of exercise in my writing seminars, especially if students are asking Chad about something they already know about. With familiar topics, this exercise might help students focus on the ways that debate happens and how they might make similar moves in their writing.

I might also ask my students to assess Chad’s posture of authority on this topic. “How does that land with you? Is this a posture you adopt in your academic writing?” The answer to the second question is, yes, all the time.

Okay, your turn. How might you use ChatGPT or similar tools as audiences for your students and their writing?