The Future of the Internet is Already Here, You Just Didn't Notice
AI Breakthroughs Reshaping Our World: From Viral Pope Images to GPT-4's Advanced Capabilities
Today’s article is going to be an update on what I’ve learned about AI in the last couple of weeks. Some spicy stuff has happened including a fake AI generated image of the Pope going viral, and improvements to ChatGPT’s AI behind the scenes that will totally destroy the internet as we know it. But also add a lot of cool stuff as well. Check out the article for more.
Dripped Out Pope Is Fake
So credit for my understanding of this was an article over on Polygon by Nicole Carpenter called Puffer coat pope is fake, but the AI art’s impact is real. In this article they credit one of the newsletters that I follow called Garbage Day by Ryan Broderick, but I am nothing if not an honest reporter and I TELL IT LIKE IT IS DAMMIT!
So for the chronically online members of my audience (Hi) there is a pretty high chance that you may have seen this image of the pope looking like he got his winter outfit by recommendation from Kanye West. Its the kind of thing that is perfectly plausible enough for people to latch on to and spread like mosquitos with malaria. Of course when this happens the context of the image is usually lost.
The image was created (rather transparently) using Midjourney a popular AI image generator that you may have heard of. It has become amazingly good in the last couple of weeks. I use it myself to generate my thumbnails! I usually throw them into photoshop for some final touches, but as a first hand user I can tell you that the quality of the photos generated are simply incredible. You probably had no idea the images I’ve been using were made by an AI.
I think the reason an image like this going viral is so groundbreaking is because, part of the way that we know when something is a success is when it becomes a ubiquitous part of our lives without us really even noticing. Like photoshop, image generators are so good at creating images now that we basically have no idea if they are real. As time goes on that line between real and fake images will continue to blur. So if you are sitting around waiting for it to happen, I’ve got news. It’s already here.
Additionally the AI that I was using to voice act my articles is now being used in many ‘educational’ TikToks that I’ve seen. I have a good ear for voices and can tell right away, but I would not be surprised if most people don’t. It’s near impossible to tell.
(I looked for an example of the AI VO but couldn’t find one. I didn’t think to save any of them when I came across them)
GPT-4 Is Going To Break Everything
The newest version of ChatGPT known as GPT-4 is able to do something incredible. Outside of being a smart chat bot, its almost capable of doing everything a human can do on the internet. Including seeing pictures. Its able to look at pictures and describe what is in them. Meaning it can solve those “Which of these is a bike” captchas that were used to test if things were robots or not.
The implications for this are massive, and again if you are wondering when you will see this tech being used the answer is it already is you likely just don’t know it. I normally wouldn’t just blanket recommend a Two-Minute-Papers video, but honestly just watch it!
The TL;DR of the video is that the AI can ‘almost’ become a lawyer, ‘almost’ become a doctor, can solve captchas, and now it will give India a run for its money. Since the AI is… a robot. It can be given a companies terms and conditions in order to be the perfect customer support agent. Additionally I have no doubt that it will be given the tools to make changes to people policies, offer refunds, and do all the other stuff customer support agents do in the coming years.
Other interesting applications for this include a program that Duo-Lingo, the language learning app, is testing by having the chat bot offer tips and role play sessions so you can practice speaking another language without needing an actual human sitting at a desk talking about how much they love croissants with you.
Will the future of the internet be great, or will all this AI enhancement destroy the human race? Let me know on twitter, instagram, or through the chat in the substack app!
That’s my AI lowdown for the week. Truly the video by Two Minute Papers shows a great image into what our future will be like so please give it a watch.
If you liked the article please consider subscribing for more! Also please consider a monthly subscription to fund my problematic spending on games and new technologies. Thank you.