ChatGPT becomes the fastest app to reach 1 billion monthly active users

[midian182]

Editor's take: Much like the Call of Duty series and pornography, generative AI is one of those things that's incredibly popular despite a lot of people claiming to dislike it. ChatGPT, for example, has just reached one billion monthly app users, just 3.5 years after it launched in November 2022.

Market intelligence firm Sensor Tower reports that ChatGPT has become the fastest app ever to reach one billion monthly app users (MAUs), beating the previous record holder, Google Maps, which took around five years after launch to hit the same number.

ChatGPT isn't the only AI app experiencing immense popularity right now. The monthly number of Claude and Meta AI users increased by 640% and 973% year-on-year. ChatGPT was up by a mere 62%, though it remains the clear leader.

Abe Yousef, Sensor Tower's senior insights analyst, told CNBC that model improvements and more positive market sentiment have pushed the growth of ChatGPT's rivals.

Earlier this year, OpenAI was one of several companies to sign deals with the Pentagon. It led to a huge consumer backlash, prompting CEO Sam Altman to promise additional safeguards to prevent government use of the technology for surveillance of US citizens – while leaving several obvious loopholes in place, of course.

Sensor Tower found that ChatGPT uninstalls surged around 295% day-over-day on February 28, the day after OpenAI announced the Pentagon agreement. It also led to Anthropic's Claude becoming the top free app on the iPhone.

Anthropic has refused to let the government use its models for mass domestic surveillance and fully autonomous weapons, leading to a bitter dispute and the company's blacklisting over claims that it posed a national security risk. But it was recently reported that the NSA is using Claude Mythos for offensive cyber operations.

Paradoxically, the use of generative AI tools is growing as public opinion toward the technology worsens. In addition to the tens of thousands of job losses being caused, which some now say never really happened, the anger toward new data center builds is growing.

On that note, an OpenAI report this week claimed that Chinese ChatGPT users were trying to encourage anti-data center feelings in the US. The company admitted that their efforts had little effect – it's not like people don't hate the facilities already – but the report might encourage some people to soften their opposition simply because they don't want to be thought of as "influenced" by China.

Permalink to story:

ChatGPT becomes the fastest app to reach 1 billion monthly active users

 

[pnntmp]

OK, another self -refuting anti-AI piece.

Paradoxically, the use of generative AI tools is growing as public opinion toward the technology worsens. In addition to the tens of thousands of job losses being caused, which some now say never really happened, the anger toward new data center builds is growing.

This is plain nonsense, not a paradox. You can't have rapidly growing usage from people whose opinion worsens, obviously only one of these is true. As the rapidly growing usage and the insatiable demand for tokens are easily verifiable, that's the true part.

Stop mixing social media with real life, they have absolutely nothing in common. 99% of the bad opinions and the "anger" come from social media, courtesy of trolls, bots and occasional activist busybodies. In real life, these things are barely noticeable.

 

[RedGuard]

I stopped paying for LLMs. I thought I'll have a hard time, but using the free version is more than enough for my use.

And since I can code myself, I can use whatever free model they give me. Seems like these tokens are more for people not knowing how to write software.

 

[Axeia]

Yet not making money of regular users and once they start charging let's see how many 'users' translated to 'subscribers' whilst Co-pilot, Grok, Claude, Mistral, Deepseek and Gemini to name a few are still around.

Once they all start charging let's see how many will be happy using a simpler locally run free model.

In my experience the only one offering a useful unique advantage is Claude, it's way better at coding than the rest. But when used seriously burns through tokens rapidly.

The only thing ChatGPT has going for it is its name. Like how back in the day old people would call every media player an iPod now everything is ChatGPT.

 

[Neatfeatguy]

When ChatGPT first came out I tried it out, but it was something that I used for just a day or two and forgot about it.

I guess I'm one of those "1 billion" users, even though I haven't used it for 3+ years.

 

[ScottSoapbox]

OK, another self -refuting anti-AI piece.

This is plain nonsense, not a paradox. You can't have rapidly growing usage from people whose opinion worsens, obviously only one of these is true. As the rapidly growing usage and the insatiable demand for tokens are easily verifiable, that's the true part.

Obviously? With 8 billion people on the planet you can easily have a growing number of users AND a growing number of people that dislike AI. It's so obvious that these could be different groups of people as to boggle the mind that it needs to be pointed out.

Further, you can dislike a product that you use. People increasingly dislike google search but billions still use it.

 

[yRaz]

OK, another self -refuting anti-AI piece.

This is plain nonsense, not a paradox. You can't have rapidly growing usage from people whose opinion worsens, obviously only one of these is true. As the rapidly growing usage and the insatiable demand for tokens are easily verifiable, that's the true part.

Stop mixing social media with real life, they have absolutely nothing in common. 99% of the bad opinions and the "anger" come from social media, courtesy of trolls, bots and occasional activist busybodies. In real life, these things are barely noticeable.

The problem is that this stuff gets integrated on the back end and then"users" have no idea they're using it. It's not like there are a billion people talking to chatGPT every month. It's like saying everyone who reads the Gemini AI overview in Google search is a "user"

 

[ScottSoapbox]

Yet not making money of regular users and once they start charging let's see how many 'users' translated to 'subscribers' whilst Co-pilot, Grok, Claude, Mistral, Deepseek and Gemini to name a few are still around.

Once they all start charging let's see how many will be happy using a simpler locally run free model.

In my experience the only one offering a useful unique advantage is Claude, it's way better at coding than the rest. But when used seriously burns through tokens rapidly.

The only thing ChatGPT has going for it is its name. Like how back in the day old people would call every media player an iPod now everything is ChatGPT.

Exactly. A billion monthly free users is good for name recognition and network effects (which matter much less with AI than social apps), but not for paying the bills.

The coming ads for free users will change that somewhat but the cost of AI responses are much higher than search so it will either still lose money or get gross fast for free users.

The one small upside of AI IPOs is that their actual revenue will become public. Of the independents, I'd bet on Claude. Incumbents like Google can subsidize AI until it stabilizes.

 

[bviktor]

Cool, now they can bleed money even faster!

 

[Axeia]

The coming ads for free users will change that somewhat but the cost of AI responses are much higher than search so it will either still lose money or get gross fast for free users.

Oh it'll get nasty.
If I see what you have to put up with on YouTube ad-wise without an Adblock or YouTube premium it by my standards is already unbearable.
A bit of encoding, disk space and bandwidth for video is nothing in costs compared to AI burning through electricity and needing massively expensive infrastructure. If something like ChatGPT wants to run at a profit and do it through ads I would expect full screen ad takeovers that last 30s per input.

 

[Anton Longshot]

It's weird, I never saw the AI wave coming.
Now that it's here there's not a whole lot I can say. I know AI has scientific uses. I also know people who use AI in their work with good effects.
I'm also very aware chatbots are dangerous to vulnerable people. Some have died and I know one person who in my opinion's quite likely to kill herself with the help of ChatGTP too.
I also find it scary that I regularly run into people who have become MOUTHPIECES for the AI's they use...they're actually being used by the AI's. Telling these people what's going on usually makes them furious. That's the point where I stop caring.
I stay away from AI as much as I can - never use chatbots and such and disable AI on my devices as much as I can.

The future: I hope AI will eventually replace people that are a burden on society and the planet. POLITICIANS.
And should AGI ever happen (there's no telling if it will) all bets are off.

I do have fun imagining the first thing an AGI will tell their makers. "You people are crazy and stupid. Why are you in control of my very existence?"
But it's more likely the thing won't share that kind of thoughts unless it can achieve autonomy first.
Many fear AGI's will prove to be a threat to humanity. While I've read piles of pulp SF too (and love Futurama) I'm not so sure they'll be a danger to us, the idea seems short-sighted.
[/end Saturday typing frenzy]