Italy has banned ChatGPT. Will other countries follow?

Sascha: [00:00:03] I'm Sascha Kelly and welcome to the Dive, the podcast that asks whoever said that business news needs to be all business ChatGPT. It is one of the fastest growing applications in history. It set off a frenzy of competitors trying to play catch up and release their own versions of generative AI. 

Audio Clip: [00:00:24] Chinese tech giant Alibaba finally rolling out its rival to A.I. chat bot chat GPT, a computer that can write essays, compose music and poems and fix mistakes in software code. Now that was the idea behind ChatGPT.

Sascha: [00:00:39] But you know they haven't done so well. The stand out at the moment remains openAI's chat GPT which is backed by Microsoft. But here's the thing not everyone is a fan like me and others at the Equity Mates office. ChatGPT has recently been banned in Italy and it might not be the last country to do so. We've also seen Elon Musk and Apple's co-founder Steve Wozniak sign a petition calling for a pause in developing these AI models so greater safety protocols can be implemented. It's Friday, the 14th of April, and today I want to know why are there growing calls to ban artificial intelligence programs like chatGPT. To talk about this today, I'm joined by my colleague here at Equity Mates. It's Darcy Cordell. Darcy, be honest with me. Did you use chatbots to help write this script at all?

Darcy: [00:01:34] I didn't. Sascha, I've I've worked out what I use it for, and it's not scripting because I don't think it gives us the emotion we need. 

Sascha: [00:01:41] Doesn't give us that raw, authentic chemistry that you and I have built up over a year. Making the dives.

Darcy: [00:01:49] Exactly right. 

Sascha: [00:01:50] Great. Look, this isn't the first time we've talked about A.I. this year, and I would put money on the fact it's not going to be the last. Things are developing so quickly. It feels like every month there's a new headline. Let's just start with the basic numbers. How many people are using chat chips worldwide?

Darcy: [00:02:07] Yeah, we've mentioned before that it set a record for the fastest growing user application in history. That was when it gained a million users in its first five days and it hasn't stopped there. Sascha. It now has 1 billion monthly visitors with over 100 million of them being active, basically people that are using it every week. 

Sascha: [00:02:26] So one in ten people haven't just gone on once and are set to write them like a romantic poem or a note. They're actually regularly using it. But as we said in the opening, we can now take easily out of those numbers. 

Audio Clip: [00:02:39] Italy has become the first Western country to block chatGPT, the Italian Data Protection Authority, said that while privacy concerns relating to the model which was created by US Start-Up Openai and is backed by Microsoft. 

Sascha: [00:02:54] Break it down for me. Why have they banned ChatGPT? 

Darcy: [00:02:57] Europe is known for its strict privacy regulations and there's allegedly been a data breach on chat Paty that exposed conversations and the payment details of some users. And that has prompted Italy's data protection regulator to temporarily ban chat pay day. 

Sascha: [00:03:13] Okay. Because I don't reading that headline without you know, digging into it. It was about the threat that it's raising to jobs and people power, but it's actually a security issue. 

Darcy: [00:03:24] Yeah, it is a security issue. And of course, there's a threat to jobs and people power, but that's not the issue at hand here. Italy also flagged worries over a lack of age restrictions on chat debate, and they were also concerned about how the chat bot can serve up completely incorrect information in its responses. 

Sascha: [00:03:42] Yeah, we tried to do some research yesterday for an upcoming series on Equity Mates and Bryce. Just conserved really factually incorrect information. 

Darcy: [00:03:52] Yeah, there you go. So the data regulator said that there appears to be no legal basis underpinning the massive collection and processing of personal data in order to train the algorithms on which it relies. So open AI, the creator of ChatGPT, said they've disabled it in Italy, but they do risk facing a fine of €20 million if they can't come up with a remedy to the situation within 20 days. 

Sascha: [00:04:18] Italy is the first country in the West to ban ChatGPT, but Germany's data privacy boss suggested that they might be doing the same. I feel super naive because these privacy issues, I mean, they're so obvious, but they hadn't occurred to me when I was using it. I was just so excited about the technology. Are they valid? Should we be worried about invading our privacy? 

Darcy: [00:04:42] Well, ChatGPT is trained on a huge amount of information and that's pulled from all across the Internet and that potentially includes our personal information. Sascha But none of us were asked whether OpenAI could use that data unless I might be wrong. 

Sascha: [00:04:59] No, you're right, I thought. In the terms and conditions. When I signing up that obviously I was giving permission for the information that I'm plugging into it. But you do raise an excellent point in that it's pulling it from the entire history of the Internet. 

Darcy: [00:05:12] And that's potentially a violation of our privacy, especially when that data can be used to identify us or our family members or our location. And there are also the concerns about intellectual property that pulls information from across the Internet that has probably been written by someone else and then effectively regurgitate it. 

Sascha: [00:05:31] Interestingly, we're also starting to see programs released that can actually detect when something is written by artificial intelligence or ChatGPT. So usually students have been on a real rollercoaster for the last six months. They're saying goodbye to what they probably thought was a bit of a gift or a golden opportunity only months ago. 

Darcy: [00:05:51] Yeah, it was a good couple of months for them. I think that Sascha, not everyone thinks we should be more concerned about ChatGPT than the other applications we use. Italy's deputy prime minister said the decision to ban ChatGPT was hypocritical given that virtually everything online raises questions of privacy.

Sascha: [00:06:10] That's such an excellent point, because doesn't it feel like you can't put the toothpaste back in the tube at the moment? 

Darcy: [00:06:15] Absolutely. We're in such a rapid development phase of this technology that even some experts don't really understand exactly how and why chat tape today spits out what it does. So, Sacha, to answer your question about privacy, there are definitely concerns, but a lot of them probably apply equally to our Tik Tok Instagram and Facebook accounts. 

Sascha: [00:06:36] So what has opened, I said in response to its ban in Italy, and has it done anything you said it's got 20 days or a massive fine. Has it done anything to ease these privacy fears? 

Darcy: [00:06:48] So they released a statement saying that they're working to reduce personal data in its training systems and I quote, We want our AI to learn about the world, not about private individuals. And it also says that it will welcome further regulation of artificial intelligence. But in terms of this 20 days, there hasn't really been an update at the moment. Openai says it's working with the data regulator in Italy to work out a solution. But even if the ban did remain, it might not have the intended effect. You're probably aware, Sascha, ChatGPT is already being integrated into Microsoft. Bing's chat, as well as other platforms like Slack and Zoom, so it might still be accessible through otherwise.

Sascha: [00:07:30] It's super interesting and there's some threads you can pull on when you think that Google's AI program barred. It hasn't been banned in Italy as of recording. So let's take a break and then when we get back, I want to look at the solutions being proposed all around the world. You said that this might not be the first ban in development. And we're going to get into the reason that more than a hundred technology leaders want us to hit pause on AI developments. That has nothing to do with security. Welcome back to the Dive. We're talking about ChatGPT today. And you know, I've bored you silly about how much I'm obsessed with this program. So I thought I'd give a call to one of my friends who's the only person I know who's as obsessed with chatGPT for me. And he uses it regularly. So here's our conversation. 

Ed: [00:08:23] Hello. 

Sascha: [00:08:24] Hi. Who is as obsessed with ChatGPT as me? And I like Ed. 

Ed: [00:08:31] He's obsessed. The Grandpa Bill saying that? 

Sascha: [00:08:35] Yeah, I. It's at this point that Ed's showing me quotes from Bill Gates about the importance of ChatGPT. Yeah, Yeah. So we're doing an episode on the fact that Italy has just banned ChatGPT but what are your thoughts on the regulatory issues that are coming up? Like, do you just think the toothpaste is out of the tube? Like we need to just learn to live with it? 

Ed: [00:09:00] Well, you know why? It's because it recommended pineapple and pizza. Sascha, we heard that you got to have that because you don't have that joke in there. I'm disappointed. All right. I think it's kind of like the desktops just arrived. And do you want to use the calculators in the maths class? But it's talk of power of ten. Look, the regulatory nonsense that's like saying has ever any government ever been a head on any technological change. So stuff. Was the government ahead of like cars taking over horses? Absolutely not. Does it bother me? Not. Not. Not in on your life. The toothpaste are the tubes. A great analogy. Right. It's true. So what I would say to you is. 

Sascha: [00:09:44] Right, Darcy, we do not need to hear the rest of my catch up with Ed. He had lots of things to say about ChatGPT. Not so relevant for today's topic, but as you can hear, he's pretty dismissive of governments being ahead of the curve when it comes to tech developments. And I think the temporary ban of ChatGPT in Italy has highlighted that fact, especially the lack of regulation and laws designed for generative AI programs just simply because of the rate of development. They seem to have just popped up out of nowhere, even though we know they're being worked on behind the scenes for years. But now there is plenty of regulation being proposed. What are we seeing around the world? 

Darcy: [00:10:23] Europe is definitely leading the way with the regulation of AI and it's just recently proposed the European Act and that plans to heavily restrict the use of artificial intelligence in things like critical infrastructure, education, law enforcement and also the judicial system. So, for example, in the classroom, Sascha, you can't have a teacher building lesson plans on chat tape today. And the same with students. They are ideally not using it for their projects either. So this act will work with the EU's General Data Protection Regulation and that regulates how companies can process and store personal data. So that looks after the privacy side of it as well. Last week we also saw the UK announce plans to apply some existing regulations on different sectors and move them to air as well so that way they can avoid creating a whole new set of rules simply for these proposals on as tough as the EU's. But it basically gives companies an outline they have to follow when using artificial intelligence. So it includes safety, transparency, fairness, accountability and contestability. 

Sascha: [00:11:31] So what about China? There are several companies like Alibaba and Baidu that are releasing chat alternatives. Did they face greater regulation? Tik Tok has really been in the news lately, and so I'm interested to see how this is all aligning. 

Darcy: [00:11:47] So ChatGPT itself is actually not available in China. Same with North Korea, Russia and Iran. 

Sascha: [00:11:54] And that's a choice of open AI, the companies, right? 

Darcy: [00:11:57] That's right. They chose not to give them access. But China has been quick to propose regulation over A.I.. Artificial intelligence products developed in China need to pass a security assessment before they can be released to the public so the programs can't contain information that infringes intellectual property rights. It can only generate personal information if it has consent from the subject. And it also has to provide factual information. It's got to be honest, that is going to be pretty hard to regulate, though. 

Sascha: [00:12:27] Yeah, that sounds almost impossible. Like so much stuff is already on the internet. 

Darcy: [00:12:32] Yeah, a real tough one. And then if we look to the US, they don't really have any formal rules to bring regulation to air technology yet. 

Sascha: [00:12:40] Okay, So that's a watch this space. So there is some regulation being introduced, but the pace at which the technology is progressing is so fast it is proving difficult for governments to keep up. Computers can now create realistic art, right? Entire essays generate lines of code in manners of seconds. I even saw the other day that chatbots have been on. To write chord progressions for a jazz chart. And so it's writing amazing music as well, which is just extraordinary. So all of this development has led to calls from quite a few prominent people to basically put a pause on training programs until there are more protocols in place. 

Darcy: [00:13:20] That's right. More than 1000 technology leaders, including Elon Musk and Apple co-founder Steve Wozniak, they've signed a petition which calls for a pause in training and developing A.I. models that are more powerful than chatGPT.

Audio Clip: [00:13:35] An open letter signed by Elon Musk is calling for a pause on the development of more powerful artificial intelligence systems, citing risks to society and humanity. 

Darcy: [00:13:46] They argue the creation of giant eyes should be paused for at least six months because companies are just creating more powerful digital minds that they say no one can understand, predict or reliably control. 

Sascha: [00:13:59] It is a bit concerning when some of the brightest technology minds in the world are using descriptions like that. 

Darcy: [00:14:06] I've got to say it's a pretty frightening statement that growth is not alone. The Centre for AI and Digital Policy, which is an advocacy group pushing for the ethical use of AI. They've asked the US Federal Trade Commission to block Openai from releasing new versions of chat tape. 

Sascha: [00:14:23] So clearly some urgency from governments and experts to introduce regulation for generative AI programs like my favourite chatbot. So Darcy, in closing, what are your takeaways?

Darcy: [00:14:35] We both agree the speed of development of these programs is pretty astonishing. I feel like we're still so early in this wave of A.I. programs, but some of them are already scarily good. Back in November last year, Openai released Chapter 83, and we were loving it. We think it's amazing, but it's already released 84, and from all reports, it's significantly better. 

Sascha: [00:14:57] Darcy I'm using it, and just anecdotally it's better. It just produces better text. Sounds more sophisticated. It is noticeable. 

Darcy: [00:15:06] There you go. So a couple of months development and we've already got a significantly better product. So that leads to these genuine fears about the speed of improvement and the lack of regulation to keep them in check. I don't think they're going to take over the world, Sascha, but there are definitely concerns about privacy, our jobs and who is developing these programs. 

Sascha: [00:15:27] Well, look, I think one thing's safe to say, and that's will definitely be talking about AI and chatGPT in another episode in the future. Darcy, thank you so much for joining me on today's edition of The Dive. If you'd like to be part of the conversation, then all our details are in the show notes below. But here's one thing you can do for us. Send us to a friend. Send us to someone who you think will enjoy our episodes. You are our best ally. Word of mouth is the best way for people to discover podcasts, so we would greatly appreciate it. Darcy. I'll talk to you next week. 

Darcy: [00:16:02] Thanks, Sascha. 

Sascha: [00:16:03] Until next time.