Claude vs ChatGPT: The AI Drama That Changed Everything This Week (AI News)

from a new platform being created by creators for creators to people saying that they would pay

for YouTube premium if AI would stop spoiling everything. Let's discuss.

Welcome to the In a nutshell podcast. This is the daily 10 minute podcast why I take out all

of the techie talk around generative AI and focus on what the everyday person, content creator

and marketer wants to hear. I've only got 10 minutes so I can only focus on one subject.

And as this Friday, yes, it's the news. This week has gone by quick. I don't know if it feels

the same for you, but this week has gone by quick the last seven days. It felt like it was only

yesterday that I was doing the news. But yes, it's been a very busy week and I think it's because

a lot of this week was about two companies. Open AI and Claude. Now, I'm not going to go through

the whole thing because I addressed it on Wednesday. So if I haven't listened to Wednesday's

episode, it really was about Claude declining to work with the government or the US government

over weaponizing Claude and chat GBT or open AI within hours of them saying no coming running

to them saying, yeah, yeah, we'll do it. And then the backlash after that. Open AI or chat GBT

got a lot of backlash. But I don't think they were expecting this, which is Claude has become

the number one free app on Apple's App Store overtaking chat GBT and Google's Gemini and downloads

following all of this controversy. So from a marketing point of view alone, Claude has done

very very well this week. Now, would that lead to money? I don't know. But it's definitely got

more downloads this week and a lot of people has moved across. As I said, right now, for me,

I kind of have moved across. I am starting to use Claude a bit more than I was before. But I think

that complexity will be my main one. I think that's going to be my main one. And I'm not going to lie,

I'm still going to use chat GBT. I'm still going to use chat GBT because, well, I spoke about it on

Wednesday. Why? But chat GBT will not quiet. They said, you know what? Let's do what we do best in

this updated system. They release chat GBT 5.3 instant, which is a newer model. And it's kind of

designed to make everyday conversations smoother. Let's say there's a better tone and answers that

actually get to the point. They say that it makes fewer mistakes or hallucinations. They say it's

down by 20%. They also say that it does a better job combining web results with its own reasoning.

Also, they feel that the responses feel more direct, relevant and natural. I actually think

the responses bit also depends on how you train it and how you want it to respond back to you. So

they say it might be that for just a general way that people would use it, either default,

but I know that my chat GBT is very blunt and to the point with me because I asked it to do it that way.

But as I said, everyone's moving over to Claude. But Claude is not, because they turn down the

government, it doesn't mean that they are holier than thou. They've got their hands dirty before.

One of the things, or one of the controversial things about Claude was them losing a copyright

case because, well, they lost it because they stole a whole bunch of copyrighted material and

made a profit from it. Now, meta, they've actually want to go down a different route. Well,

I'm using quote marks when I say different route, but they've signed a multi-million dollar AI

license in Delwood news club, my old employer. So this deal is allowing it to use content from the

Wall Street Journal and other titles in chatbot responses and to train its AI models.

The deal was reportedly worth up to $50 million a year over a free year period and covers multiple

brands across the US and UK. The crazy thing is is that what's happening is that somebody called it

a woo and sue strategy. So they say, and I want to work with you, but I'm suing you at the same time.

It feels like 20 years ago when I was at Myspace and Universal. Universal was suing us, but they

wanted to work with us. I'm not even Universal, all of the record labels, suing us but wanting to

work with us at the same time. So it feels like the same thing is happening because most things

just go in circles. It just repeats itself, but at least meta are the same. We're trying to do it

the natural way. The issue I have with this, though, is this is where I start seeing things like,

when I ask a question that answers on news court publications will, what wouldn't be

surprising if they all of a sudden get to the top of your searches? Hmm, wouldn't surprise me.

Now, I do this every Friday, but if you want to stay up to date with the news daily, I have a

WhatsApp broadcast list. So what I do is I post three or four news articles every day in this list

just to keep people up to date with pretty much what's going on for content creators, marketers,

and everyday person. So if you want to join, go to my website, which is Andrew Miles Davis,

Davis belt, D-A-V-I-S, AndrewmarsDavis.com at the menu bar at the top, go to resources,

and under that it will say WhatsApp, and that's how you can sign up.

Now, there's a new platform by creators for creators that is USP is to root out AI deepfakes.

So creators Zander Small and Morgpy have launched FanLoc, and a point of FanLoc is to detect and

remove AI deepfakes across more than four million websites. So FanLoc, it targets gaps left by

existing DMCA and leak removal services, including coverage of Telegram, which has become a major hub

for AI generated deepfake abuse. It looks like it's trying to tackle a problem now, and it's

going to be a larger problem moving forward. So let's see what happens here, be interested to see,

again, if I find out more about this, FanLoc, I'll keep you up to date.

Now, a place where there is a lot of deepfakes is YouTube, and recently there was an article

that I read, and it said, I'll pay for YouTube Premium if it stops spoiling everything. I actually

pay for YouTube Premium anyway, but what this article is about is saying that YouTube's recommendation

engine is brilliant at discovery, but terrible at avoiding spoilers. With FanLoc's auto-players

and shorts, routinely ruining films, shows, games, sports, and more for users. Completely agree.

I hate this life. Let's say, for example, there's a football match I wanted to watch, or a wrestling

event that I didn't get to see, or a film, I have to either don't go on YouTube until I've seen

it, or if it's a film, especially the Marvel films, I'd have to go to the cinema the day it came

out. Not because I wanted to see it desperately is because I didn't want somebody to spoil it.

So, people are saying that YouTube already has the AI capabilities to power a paid spoiler filter

on YouTube Premium, so it can use its content analysis to hide, blur, warn, any spoiler heavy

videos around major releases. I wish they did this. I pay for YouTube Premium, and that will be

another thing that will keep me paying for it. Definitely keep my eye on that one. That's more

of a personal one as well, like I really hope they do that. So, let's see.

Now, an area that I haven't really done many episodes on this particular industry, but it's

a fashion industry, because AI adoption in this fashion industry is moving faster than the market

realizes. So, UBS found that 96% of the 45 major fashion and luxury companies are already using AI.

So, there's far ahead of what most investors assumed. Brands such as Ralph Lorraine are quietly

using AI to cut discounting, improve inventory, and personalize marketing, which UBS says is

already feeding through to margins and returns. What I thought was interesting was that UBS argues

that market is over-focusing on distribution risks and underpricing the near-term profit upside,

which ultimately suggests fashion and luxury valuations could be re-rated as AI's tangible value

becomes more visible. I want to help them do that. That would be quite interesting to see how they

go about tackling. Now, the last bit in news I want to talk about, well, it's more of a statistic,

and that is Gen AI was 2025's fastest rising app category, but social media still left the field.

So, social media added 108 billion hours of global mobile app time in 2025, which is over three times

more than Gen AI and AI assistant apps. Time spent on AI assistant apps jumped 425% year over year,

while in-app purchase revenues in Gen AI assistants climbed 254%. And 2025 was the year that everything

really got started from the mainstream, I would say. So, imagine what it's going to be like this

year. Will it overtake social media? I can't see it because a lot of what we do in AI is connected

to social media, but I can definitely see these numbers increasing over the years. So, in a nutshell,

that was the news this week.