(Alene) Joining me now to share how Florida International University is preparing the next generation of journalists for the world of AI is Doctor Susan Jacobson, director of the Lee Kaplan School of Journalism and Media at FIU.

Welcome.

Thank you for being here.

(Susan) Thank you for having me.

(Arlene) So we all know that AI is emerging, and now schools have the challenge of implementing this in their curriculums and policies.

Tell us how that's going.

There’s kind of two groups of AI.

There's the generative and there's the assistive.

And right now we're having problems with the generative AI, AI that actually creates content.

And the place where we're seeing it and trying to figure out how to deal with it is where large language models can actually write papers and passages for student papers, and sometimes even just automated tools like Grammarly that they've been using for a long time, will now rewrite entire passages.

And when you're talking about students taking a class on learning how to write or they're being evaluated for their writing, you know, that's just not something that we can accept.

We’re learning now, just now how to how to deal with that.

(Arlene) This is a learning experience and a learning curve for professors and students.

Tell us how you tackle that part of AI.

(Susan) Okay so from the student perspective, our students, you know, they have busy lives, they have families, they have jobs, they have school on top of that a lot of times.

So sometimes it's very tempting to use a shortcut like a ChatGPT to write your paper.

And we understand that pressure, but they really should not be able to do that.

And they don't understand also that these large language models, they're terrible writers.

So they would be so much better writing their own copy instead of using the ChatGPT because their own voice is much better, much more authentic.

ChatGPT does things like it makes stuff up, it gives you fake references, you know, it's just not a good writing tool if you just use it, you know, do a copy paste.

And what we have at FIU is we're using some AI detectors.

They're not perfect, but they will show us.

They'll give a score to a student paper as to, you know, what's the likelihood that all or most of it was written by an AI.

And so we use that as a guide.

So generally speaking, we know that if the student has kind of a low AI score, then we're not going to worry about it.

But if they have a high one like close to 100, then, you know, we have to have a conversation with that student.

(Arlene) There are some advantages, though, to implementing this type of technology.

Tell us about that.

(Susan) So in in journalism and media there's a few things that are really great for our students.

One is any kind of AI tool that helps you with video or audio editing.

In the area of video editing, for example, Adobe Premiere is a great video editor that has incorporated some AI tools so you can literally shoot your own video, bring your video into this program, it generates a transcript, and you can edit the edit your video by editing the transcript.

Now, you may need to go back and make some adjustments and move things around.

But as a first pass editing tool, it's fantastic And I grew up, um, editing tape for video, and I saw the transition from there to to digital, which is much easier.

And now this from, you know, to just be able to use the transcript, it's a leap in improvement.

There's a similar there's similar products for audio.

There's a product called descript that will allow podcasters, for example, to take their very long, um, audio sessions and, you know, edit them pretty quickly using the transcript.

Those tools are fantastic, and I encourage students to use them.

Um, the AI tools have also been used to generate code for websites to do some little programming, web based programming, you know, it'll help generate some code for you.

I think all of that is is perfectly okay.

The image editor I is the still image editors.

There kind of a mixed bag.

Sometimes they're okay to use because like if you're doing a flyer, for example, or a social media post and you need a graphic, they can sometimes generate a perfectly okay image.

The situation there is you just have to look at it carefully, you know, to make sure there's no obvious errors like extra fingers or misspelled words.

(Arlene) You know.

There is that idea where people do have these deep fakes that we've seen.

That is I feel a danger of I and I'm sure as a, you know, someone who leads a college in this world, we have to be careful with that.

What are your thoughts on deep fakes and the negative aspects of AI?

(Susan) So specifically talking about deep fakes.

One thing I can say is that the video deepfakes like if you've seen there's been some examples of Obama, you know, saying crazy things, the video of a real person.

It's not that great at this point.

It's going to get better though, right?

I think it kind of falls back to sort of some basic media literacy principles of being able to check your sources.

Where did this come from?

If it seems like it's fake, it might be fake, especially if it makes you angry or makes you sad.

Or you go like, oh, what is that about?

I can't believe that that's always a cue for you to check, to make sure that you know, what you're looking at is legitimate.

(Arlene) And journalism students, what's their feedback?

How do they feel?

Is it a mixed bag?

I know when I say it's a mixed bag, for me.

(Susan) It is a mixed bag.

I would say again, the editing tools are really helpful.

The video editing tools, the audio editing tools, um, the other tools, again, students are kind of naive that they're not completely sure of, you know, what the limitations of these tools are at this point?

Because we're early in the game, I think they will become more savvy as time goes by, you know, and they'll be able to use them more wisely.

Um, and, and part of what we plan to do is to figure out how to integrate good use of AI tools, especially things for data analysis, for doing some content generation under specific circumstances, like writing a summary, um, under specific, but in, in a specific circumstance.

And part of since they're journalism students, it's very important to excuse me to be very transparent with your audience about, you know, what the source of your material is.

If it is generated by an AI, you want to let your audience know that.

(Arlene) Tell us about some other cool technologically forward things happening at FIU.

The immersive studio for Altered Reality.

That sounds so.

Cool.

(Susan) Yes.

The immersive studio for Altered Reality is our immersive media facility.

We just built it a couple years ago, and we have students in our new game and animation program that use it to create their projects.

We have what's called a volumetric studio.

And what this is, if you're familiar with Pokemon Go, you know how you can see Pikachu in your in your environment?

Well, the Volumetric studio is basically a series of cameras in a circle that creates a 3D video.

So it could be a person, it could be an object.

And then you can take that video file.

We call it an asset.

And you can you can see it like you would see a Pokemon Go character.

You can insert it into an immersive environment like Unity or Unreal, which is the platform that they use for games.

You know, you can use it in films and animation, but and it's a complete 3D character.

So, so yeah, so we're working on that and we're, we're trying to figure out how to integrate that into the journalism business.

Um, we haven't gotten there yet.

I'm trying to.

Think of the ways.

But we're working a second.

I'll get back to.

You on that.

Yes.

Data visualization, for example.

Yes.

For sure.

Yeah.

(Arlene) You have the big networks.

They show all the data and will be pointing at it.

So I can see maybe something happening with that, making it more compelling and interesting to take in a lot of data.

(Susan) Yes.

And TV.

(Arlene) Usually.

(Susan) TV networks are using what's called virtual sets.

And they use these technologies.

They'll use Unity or Unreal, which again is like a game platform to create sets that aren't really there but really look like they're there.

(Arlene) It's great that the students are on the forefront of all this and they're learning this.

But can journalism survive?

AI (Susan) I absolutely believe that journalism will survive.

AI there's no there's no substitute for actual on the ground reporting.

One of the things that we are seeing, however, is increasingly, if you ask anyone under the age of 40, surveys show that a lot of them say that they get their news from social media, and they're not following the New York Times on Instagram.

Often they're following what we call informal information providers.

These are influencers and podcasters, you know, who often have a topic that they talk about on a regular basis with their audience, and they're not necessarily trained in the norms of journalism.

One of our missions is to help these informal information providers understand better the norms of journalism and how to be good information providers, and also to prepare, prepare our journalism students to be these kinds of informal information providers.

(Arlene) Giving news first and then learning how to do news second.

That's so crazy to me.

But the responsible implementation of AI, I think, is the key.

Yes.

Thank you so much for joining us, Doctor Jacobson.

(Susan) My pleasure.

Thank you for having me.

(Arlene) Absolutely.