Jon Rubin is considered by many in the virtual exchange/COIL world as the father of this tech-based pedagogy. He’s been involved in COIL since 2001 and he helped start the SUNY COIL Center in 2006.
During the last five years, Jon has been active at COIL Connect which supports the international growth of COIL Virtual Exchange.
He keeps up with the latest technology in the field and right now he is very focused on AI both in terms of its global impact and its likely impact on COIL/virtual exchange.
AI in a global context
“I’m very involved with technology and change. But technology has a built-in bias. Every technology has a bias. And it’s often invisible to us. I first realized this when I read a book over 40 years ago which really stayed with me. It was called: Four Arguments for the Elimination of Television.
Jon explains how it appeared far-fetched back in the 1970s and how people laughed at it rather than take it seriously.
“But’, says Jon, ‘the author made a very good case why TV would have a negative impact on society and on people.”
The author advocated for some control in terms of what TV portrayed, as he thought that it was much better at portraying violence than it is at presenting peace.”
Although he was referencing the TV of the 70s which were, in fact, small boxes, it turns out there was more than a hint of truth in that assessment.
Jon says this idea of technology kind of having a ‘mind of its own’ has stayed with him ever since. And it is not necessarily dependent on the people who invented that technology.
The influence of rich white men
Jon Rubin says that we have some very rich men who have invested phenomenal sums in the hope that AI will be successful. Ultimately, this means it’s a technology that cannot lose.
“It has to win. Even if all of us said, this is a terrible idea – it would be hard – because the amount of money being spent on promoting it and developing it is huge.
And then, finally, we’re living in an age where lies, untruths, lack of veracity and authenticity has become almost the norm.
I mention the President of the United States, who’s the most notable person in this regard, but many, many others, will literally state to the public things that we know very well are not true. But people will still believe him because of the psychological force of the divisions in the world. So it’s very easy now to lie. Lying is probably the main modality of speech in the present.”
Jon is concerned that AI, whether used for the news, or for virtual exchange will serve to encourage this untruth.
“If that untruth pervades our virtual exchanges, then virtual exchange is in trouble.”
Some positives of AI on VE (but not many)
Jon agrees with Simone and Mark that there are some positives about AI tech on VE.
He likes how it’s possible to have easy transcripts and summaries following zoom meetings which include action points for example. It means it’s not necessary to view the entire recording to get the required information you need.
And given that VE depends also on speaking different languages, its translation capabilities are a huge advantage in terms of leveling the playing field.
And more negatives……
Jon thinks this will ultimately turn out negatively, though.
“I have to make one technical point first. Presently, most use of AI is not done in real time due to the amount of energy needed.
Meaning, let’s say somebody is creating a deepfake video.
They put together their scenario, and the AI generates that video.
This is a simplified example: let’s say you ask the AI to show an immigrant jumping out from behind this car and shooting an ICE officer.
And then you wait a couple minutes, and then that video exists.
Whether it looks truthful or not, it’ll probably require a lot of tweaking. But eventually, you create something that the average viewer could not tell is not real. This still takes time.”
However, Jon thinks that we are within a year of being able to make these things in real time.
“So let me give an example. This is an extreme case. Let’s say you have a call with Zoom as a central tool between the UK and Korea. But one of the Korean students is very shy or feels their English is bad.
So, they bring in an AI tool, a real-time AI tool, and they type in: ‘Make me seem courageous and arrogant.’
Now, when they talk, the AI translates in real time, but it’s taking a different tone of voice. It’s not speaking like this student.
And similarly, to carry it to the ultimate – the person says, I actually would rather look like a Caucasian.
And instead of looking like I’m 18, make me look 25, because I’d rather look older.”
Although an extreme example, Jon thinks we are not far off from this being possible.
The impact of AI on COIL/Virtual Exchange
And this is the heart of the problem according to Jon. Participants in a COIL will no longer be able to distinguish who they are talking to or whether they are real. This is how trust is eroded and virtual exchange will no longer be viable.
“We will no longer be able to tell if the people we’re talking to. It could be a 60-year-old man who’s told the AI to make them look 22.
It could be a woman wanting to look like a man. This is all around the corner. It’s not sci-fi anymore.”
Despite the small positives that AI does bring us, Jon says we’re close to a place where, unless we set some boundaries, we’re going to head down this road, “…and I feel, because it’s my field, that COIL and virtual exchange are going to be very, very vulnerable.”
Right now, the VE format says it’s possible to have a valid, meaningful, and moving international exchange through online means. But we could be close to a point when people no longer trust who or what they see in front of them. They will need to be sitting next to that person and shake their hand to believe it’s a person.
Personal experience
Recently Jon says a strange message came through his Facebook channel which he found a little odd. He also says he rarely responds to people he doesn’t know, but for some reason, this time, he did.
“And after about 15 minutes, I realized this wasn’t a person. It was a chatbot. It was really good.
And of course, it was well armed with photographs and such. So, after a certain point, when I asked it some really peculiar and difficult questions, its answers came back very confused.”
At this point, Jon cuts contact as he realises it’s not real despite being terribly convincing, with a history, and an identity, and a whole background story. He doesn’t know why he was contacted. But he suspects he would have been asked to purchase something like bitcoin or cryptocurrency.
Time to fight back
“I’m not a revolutionary. I’m not saying join my army. That’s not at that point. I’m trying to raise awareness.”
I think there are places where people can fight. What happened in Minnesota a month ago, which I’m contextualizing in U.S. politics, but indeed, people stood up, people defended. And eventually ICE more or less left Minneapolis. Not totally, because it’s a corrupt government that’s pushing this agenda. But people push back. And I think you can push back on anything. It’s just that AI is more ubiquitous.”
Jon says it will be a tough battle, and he doesn’t have a battle plan, but he wants us to understand where we are and where we are going, before we end up having to memorise books like in the final pages of Fahrenheit 451 when people are pacing the forest trying to memorise books before they are lost forever.
A dystopian finale for sure!