I tried using AI. It scared me.

All I wanted to do was fix my email. I didn't expect to have a minor existential crisis about how much the world might be about to change. Normally, this is the kind of talk I would have given on stage at some conference about the future of the web, but, well, I'm in New Zealand right now and I also have a sneaking suspicion that those conferences might not be relevant because more time. I was lucky to experience the mass adoption of the Internet. Anyone under 25 who sees this won't remember how quick and incredible that transformation was. And I was at the forefront, I was living it, I was one of the people who understood what was happening.

Useful new technologies generally follow a sigmoid curve. There's a slow takeoff as the technology is invented and the major pain points are fixed, and then you have this incredible explosion of growth as people find it useful and the technology gets better and better and better and better and better and better. many competitors enter. the market, and people are finding more and more practical uses for it in the real world. And there's this massive rush to make newer, better, bigger things that people keep making more things with. And then the limit of what is possible with that technology is reached and the pace of progress stabilizes again.

Basically, there's a reason why very few people camp outside Apple stores to buy the new iPhone anymore. I remember Napster back in 1999. And looking back, I think Napster was the first big sign of how many industries were going to be changed or completely destroyed by the Internet. Not just the music industry, but also travel agencies, video rental stores, encyclopedias, shopping malls, department stores, the postal service, journalism and the entire media industry. Napster wasn't the first signal, but it was the first major signal that millions of non-technical people used by choice. The first to receive great media attention and legal problems.

It was the first great popular disturbance, the first warning shot, the first thunder in the distance that said that the storm was approaching, that everything was about to change. I don't think many people realized that at the time. And even if they had, they couldn't have predicted the world that would come next. The world I grew up in, the world I feel really comfortable with, the world of the web, of social media, of smartphones. This video you are watching now is a video from early 2023 and is very much written in English. This is a record of a point in time and space.

And I may be wrong about this: I don't know where technology will go from here. But: I think that world, my world, that I grew up in, is about to change radically. I think we are on a new sigmoid curve and I have no idea how far along that curve we are right now. And I don't know if I want to change with that... but I think I'll have to. And all I was trying to do was fix my email. This is a really specific problem, but it really bothered me and I have to explain it so the rest of this makes sense, so bear with me.

Almost all email systems have the idea of ​​folders. You have a copy of an email, and once you're done with it, you put it in a folder like a piece of paper in a filing cabinet, because that's the mid-20th century physical analogy that inventors followed. But. In 2004, Google launched Gmail, which ditched all that in favor of labels. Email is in a big pile, you look for things and add tags to help you sort them. I hate labels, because I like the folder analogy, that's what I started with, that's what I was already

using

. I didn't want to change for no reason.

But everything else about Gmail was so much better than any of the alternatives at the time, so I switched to it a couple of years later. And I decided, well, I'll treat tags like folders and change my workflow as little as possible. Each thread receives a label. Is that like putting it in a folder? Which meant that when I backed up my emails, which I do, I never rely on having just one copy of something, especially if that copy is in the cloud, but when I did that, I could use an email program old school desktop software like Microsoft Outlook to make that backup.

You can use programs like that with Gmail accounts, they communicate with each other. If they didn't, things like the default mail app on iPhones wouldn't work and people would complain. But those old school systems don't understand labels, so Gmail just lies and says they're folders. Excellent! That means my email program can contact Gmail and request a copy of all those messages, one "folder" at a time, and then I know it has a backup of every message. They are even stored in real email folders, like I wanted! It took me fifteen years to realize: that wasn't working correctly.

I was doing a routine backup last month and noticed that a thread in that email backup was missing several recent messages. And that made me suspicious, that kind of discomfort you feel when you know something is wrong, but you still don't know what it is. So I searched some more and realized that a lot of messages were missing from my backup. More than a hundred thousand of them, in fifteen years. They were still there in Gmail, on the web. They had not been lost. It's just that they hadn't been supported. And that's because, in Gmail, labels are only attached to individual messages and not to the entire thread.

That is invisible to most people. It is barely mentioned in the documentation. And it's a really subtle and annoying distinction. If you send an email, put a tag on that thread, get a reply, and then just hit "Archive" - ​​that reply doesn't get the tag. And as you go and go and add content to that thread, none of the other messages there get a tag unless you explicitly untag and retag the entire thread. Which means that over 100,000 untagged messages were not in my backup. Is this a really specific, nerdy problem that only bothers me? Yes absolutely.

I am one of the few people in the world who will be affected by this. Many people don't classify or label their emails at all, and most don't back them up. But I needed to fix this problem, I needed my backup to work and I didn't want to change my damn workflow. But how can I fix all the email threads for fifteen years, without having to painstakingly click through all of them? I grew up fixing things with code, so obviously that's where my brain went. Google has a service called Apps Script. It means anyone can write simple code to easily change and automate things in Gmail, Calendar, Docs, YouTube, whatever.

It's a brilliant service, I have no idea how it was done! And inevitably, one day, it will shut down without warning, because that's precisely what Google does. But Apps Script meant that to fix this problem, I could spend about an hour writing some code. And it's been a while since I did that! The first thing I do when I'm coding something this simple is write down what it needs to do in plain English. Just the logic flow so I can use it as a guide. So I wrote it and then I thought, wait, isn't there an AI that's meant to do this now?

Yes, this is a video partly about ChatGPT, but I promise I won't go into details there. Telling someone about your fascinating conversation about AI is like telling someone about your dreams. They don't care, it just sounds like you're hallucinating nonsense. But I had seen posts from people freaking out about how good ChatGPT is, and intellectually, yes, I looked at it and agreed with them. It seemed surprisingly good. But it's just a text transformer. All you do is guess what the next word in the sentence is. I knew we'd seen some improvements in that lately. I made a video two years ago about “sentences that computers can't understand, but humans can,” and at the end of that video, I made a joke about how computers with language skills are ten years away, just as I said. have been. for the last 40 years.

I made a mistake. Turns out that's pretty much resolved now. But ChatGPT is still just something that predicts what the next word will be. It produces bland, generic results that are almost certainly wrong about many things. It will help spammers and maybe have some real world use. Good? But I thought, okay, let's put in my text description, tell it to translate into this obscure Google Apps Script, and see what it does. …and wrote the code for me. In some seconds. And I felt this existential horror in the pit of my stomach, and at the time I couldn't explain why, but I found out later.

It is simply predicting the next word. The code it returned to me was a little off, but frankly my first attempt would have been a little wrong. So I asked him in plain English to correct a couple of errors he had spotted, and he did, and that version was almost correct. Although not completely. He used a really strange approach. I

tried

to figure out why he had done that and then I realized I could ask. And he told me, in English, why he had done that. And he had made the same mistake I did with my backup.

As Google's documentation turns out to be incorrect, the Google Apps Script itself treats tags as if they were attached to threads, not messages. Google misspelled its own programming language! So I told ChatGPT about the error. And he fixed the script. It worked its way through Google's bad documentation to build what I was actually asking for. And after all the back and forth, I didn't really save much time! But it was much easier. I never had to go through documentation to find the exact specific magical incantation I needed. Also, my coding skills have been outdated for years.

That's one of the reasons I don't make computer science videos anymore. The last time I learned something new in code was probably in 2015. My old skills were good enough. I didn't want to change my workflow. I'm not sure if ChatGPT is a better coder than me, but I feel the same way. I felt like I was watching while someone wrote code, and every once in a while I would chime in and say, "um, I think you might have made a mistake there?" And sometimes I was the one who made the mistake. At one point, I realized that I had written something wrong in that first text description, ChatGPT had simply been following my incorrect instructions.

I know it's just a text model that predicts what word comes next, but it didn't seem like it. So at first, I thought that feeling of existential dread in the pit of my stomach was because my brain had said "that's a human" or "that's an alien intelligence." Or because I was starting to think, well, what if my brain is just a transformative system trying to predict what the next word will be? I'm sure I remember some old popular science article about how the brain is a prediction engine and "surprise" is just a fancy word for "being wrong about what comes next." But then the next day I gave ChatGPT a harder problem, one that required not only translating English into code, but also a small leap of logic, and it completely and confidently failed in multiple, obvious ways.

And that meant I didn't worry about him coming for my job. Not yet, anyway. Of course, no one had asked before that exactly that problem, that of labels and messages, be solved. ChatGPT was not just about copying and pasting from somewhere, but possibly involved creativity. But if I'm honest, a lot of the programming I used to do was just copying and pasting bits from somewhere. At some point, every modern programmer searched Google or Stack Overflow and "adapted" something someone else wrote. It's basically part of the job at this point. By the way, I'm deliberately avoiding the argument about creativity, the art of AI and copyright.

I don't know how I feel about it and I don't know how the courts will resolve it. If you hired a human artist or musician to create a painting or song in the style of a more famous artist, it would be strange... but it would be legal. As long as they don't claim to be the famous artist, as long as the smaller artist doesn't market their name, it's fine. The style of most of my videos on this channel is just a slightly faster version of the TV shows I watched as a kid. That is what this intelligence was trained in.

And this video is basically the style of any other video essayist on YouTube. It's just that I tend to go to the beach more frequently. If someone set up an AI to generate YouTube videos, or even educational YouTube videos as a genre, do I think you'd be okay with that? I don't think I have any problem with my work being a small part of a massive data setof training. But what if someone started specifically scamming my name and my job? Yeah, I can see why artists are furious about that. The British legal term for this is "impersonating".

I've said it before: please don't train models with what I do. Or, frankly, against someone specific without her consent? But hey, I wasn't worried about being replaced, and I wasn't worried that I was talking to some alien artificial intelligence equivalent to a human being. So why did I still have that feeling of dread? Artificial intelligence, text transformers and diffusion models, everything we are seeing today, seems to be on that sigmoid curve of progress. And I don't know what point on that curve we have reached. If we're already around most of that curve, then great. Programmers and artists have completely new tools, but they cannot create something on a human level with them.

Not much work will be needed. It will make people's jobs more efficient, in the same way that many inventions, such as Photoshop, have done before. If we're in the middle of that curve, then wow, we'll have some really awesome new tools very soon, which still need some humans to use them. Maybe Siri and Google Assistant will become what they were always promised to be. But that feeling of dread arose from the idea that ChatGPT and new art AI systems could be to my world what Napster was to the late '90s. The herald, the first big warning that this new technology, the thing that was going to change everything, was starting to change everything.

Where a large number of people, not just nerds, actively used it. And it didn't matter that Napster was sued and ceased to exist, because by then there were Morpheus, Grokster, LimeWire and KaZaA. And then there was Spotify. The old business model, the idea of ​​buying a copy of music, had taken a mortal blow and no one noticed for a while. If you hear echoes of the Napster case in lawsuits against artistic AI programs... so do I. The other week I was getting a haircut and, without being asked, the hairdresser started talking about ChatGPT and how he had used it to write a formal email that he didn't bother to write himself.

He wasn't a particularly technical person, he wasn't a nerd like me, but he used it easily, in the same way that someone who wanted to get a ton of free music would discover Napster. It's not about ChatGPT, not specifically. It's about what it represents. Because if we're still at the beginning of the AI ​​curve, if we're at Napster point, then everything is about to change, as fast and as strange as it did in the early 2000s, perhaps beyond it all. recognition. …and this time I'm on the wrong side. I'm like the music executive, back in '99.

To me, I feel like something could have gone very wrong in the now comfortable world I grew up in and settled into. That's where the fear came from. The worry that I suddenly don't know what's coming next. No one does. I've been complaining for years that it seems like nothing has really changed since smartphones came out... and I think maybe, maybe, I should have been careful what I wished for. At some point I will watch this video and in retrospect I will be able to easily see where we were on the curve. We all will. Part of me hopes I'm completely wrong, that in a few years I'll still be working like this.

The email thing is a metaphor, of course, they won't be email folders for you. It could be something else, something you are attached to. Maybe something minor, or maybe the entire industry you work in. But right now, that feeling, that growing horror, that dread: it turns out it was the worry that after years of being fairly stable and comfortable, my world is about to change. …and despite everything, I will still want my email to be in folders. …and now, an AI-generated NordVPN ad! Today I want to talk about one of my favorite tools to stay connected while traveling or just browsing the Internet.

It's called NordVPN and it's been a game changer for me. With NordVPN, I can access content from anywhere in the world, no matter where I am. Whether I'm in a country with strict internet censorship, or want to access a website that blocks people from a particular country, or just want to catch up on my favorite shows from home, NordVPN has me covered. Some other amazing features of NordVPN include the ability to connect up to six devices simultaneously and 24/7 customer support. Whether I'm

using

my laptop, tablet, or phone, NordVPN makes it easy to stay connected and secure.

So if you're looking for a VPN that can help you access content from around the world and stay safe online, try NordVPN. Trust me, you won't regret it! For the best deal they currently offer, visit nordvpn.com/tomscott or click the link in the description. Soon I will be out of work.