Business ChatGPT rolls out Ads for Free Users - OpenAI has begun introducing ads inside ChatGPT for free users, marking a shift toward sustainable monetization for AI platforms.

  • 🏰 The Fediverse is up. If you know, you know.
  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
1770737851982.png

OpenAI has started rolling out advertisements inside ChatGPT for users on the free tier, marking an important step in how the company plans to fund one of the world’s most widely used AI services.

The ads appear as clearly labeled sponsored content placed separately from ChatGPT responses. The company’s paid tiers, including Plus, Pro, Enterprise, and ad-free usage-limited plans, continue to operate without advertising.

The rollout reflects a broader shift across the artificial intelligence industry as companies move from rapid user growth toward building sustainable business models.

Rising costs behind AI growth​

Since ChatGPT’s public launch in 2022, demand for AI assistants has expanded quickly across consumers, businesses, and developers. That growth has also increased the cost of running large AI systems, which require significant computing infrastructure and energy.

Advertising is a familiar solution in the technology industry. Search engines, social media platforms, and streaming services all introduced ad-supported versions after reaching large audiences. AI platforms now appear to be moving in the same direction.

OpenAI has said sponsored content will remain clearly separated from AI responses and will not influence how answers are generated. Users can hide or report ads and adjust preferences through ad settings.

The company has not indicated plans to expand ads beyond the free tier.

A new phase for AI platforms​

The introduction of advertising inside ChatGPT highlights how AI assistants are evolving from experimental tools into large-scale digital platforms.

Operating large language models requires ongoing investment in data centers, specialized chips, and software infrastructure. As adoption grows, companies are increasingly combining multiple revenue sources, including subscriptions, enterprise services, developer tools, and advertising.

Industry analysts say hybrid monetization models are likely to become standard across AI platforms over the next few years.

For users, the immediate impact is limited. Free users may see occasional sponsored content, while paid subscribers continue to use ChatGPT without ads.

The longer-term significance lies in what the change represents. AI assistants are becoming part of everyday computing, and companies building them are beginning to adopt the same economic models that shaped earlier internet platforms.

OpenAI’s move suggests conversational AI is entering a more mature phase, where scale, reliability, and revenue must grow together.

https://techputs.com/chatgpt-ads-free-users-openai-revenue-model/ (Archive)
 
Click to expand...
This is as close as they'll get to fully admitting that they're losing a ton of money on this crap.

I'm just waiting until they start inserting brand names into the chatbot answers as a form of covert advertising.
Or maybe they're already doing it.
 
Idiotron said:
This is as close as they'll get to fully admitting that they're losing a ton of money on this crap.
I'm just waiting until they start inserting brand names into the chatbot answers as a form of covert advertising.
Click to expand...
I saw someone summarize it this way: they offered LLM ChatBots as this miracle of deep insight that had to remain pure of any commercial influence, only to have a gorillion people a day melt a million GPUs asking "Which TV best?", with 0.01% of their user base paying for premium memberships to cover the cost.

They could just be charging for ads every time they list a product, and be much more self-sufficient. If they want to keep the LLM itself unbiased, they can have it generate answers like it is now, then a separate dumb layer looks at its text and inserts links where relevant.
 
Last edited:
Reminder that OpenAI has not yet made ANY prodit, and it need to be putting out something like 20 billion a year in profit by 2030 to pay off the deals it has made.

This bubble so retarded.
 
And this is how it starts. They roll out ads for the free tier, then slowly start slipping them into paid tiers as well.

Anyone 'member when Hulu first started? Ads for free users, but you could pay something like $10 to have no ads. Then the paid users got ads. Then the price went up. And now it's like $50-$80/month and you get just as many ads as cable (if not more).

Hell, cable itself for that matter. The whole point in the beginning was no ads because you were paying for the service. We see how well that turned out.

Total. Advertiser. Death.
 
Idiotron said:
This is as close as they'll get to fully admitting that they're losing a ton of money on this crap.

I'm just waiting until they start inserting brand names into the chatbot answers as a form of covert advertising.
Or maybe they're already doing it.
Click to expand...
Simple. Just ask the bot "Coke or Pepsi?"

Whichever answer it gives is who is paying for advertisements.
 
It reminds me of the VR fad for gaming. It's amazing and revolutionary but it isn't really in the reach of enough users or seen as valuable by enough people to be profitable.

The dream of AI outpaced the uses for it.
 
"step toward sustainable monetization" is the slimiest fucking descriptor I've seen for adding the same shit every other site's currently doing. Maybe stop spilling spaghetticode and buying datacenters to house said spaghetticode for a bot you keep lobotomizing so it doesn't do anything but repeat positives back at the person that's messing with it? Surely selling your users dada to advertisers like you niggers are should be enough?
 
>tfw you're fucking your ultra realistic OpenAI sexbot in the year 2040 and after you coom it transmits unskippable adverts for prescription medication directly into your brain before releasing your dick from its vice-like robogrip
I knew I should have paid the extra 30 credits a month for the premium version.
 
From what I have heard, Ads won't be sent to free users that are under 18. So I guess if you didn't want ads and there isn't much of an restriction regarding prompts for under 18s than you have a loophole where you can sign up for an account and set the DOB to like 8-10 years old to be free of the ads.
 
JohnnyG said:
I saw elsewhere someone summarize it as: they offered LLM ChatBots as this miracle of deep insight that had to remain pure of any commercial influence, only to have a gorillion people a day melt a million GPUs asking "Which TV best?", with the 0.01% of their user base paying for premium memberships trying to cover the cost.

They could just be charging for ads every time they list a product, and be much more self-sufficient. If they want to keep the LLM itself unbiased, they can have it generate answers like it is now, then a separate dumb layer looks at its text and inserts links where relevant.
Click to expand...
If you think that's how it would stay, I've got a bridge to sell you. Google Search became almost unusable overnight because they decided it wasn't making profit fast enough. Never mind that it was practically printing money, they needed more money and faster.
A solution like that is only going to last so long as somebody who has enough authority can hold off further incursion, and that won't be long when Walmart starts getting upset that they aren't the first answer when somebody asks what supermarket is the best, and they start threatening to pull their ads, or somebody with incredibly deep pockets starts wanting to push their thing without the users knowing. It will never last.
 
Gottahavemysips said:
If you think that's how it would stay, I've got a bridge to sell you.
that is only going to last so long as somebody who has enough authority can hold off further incursion
Click to expand...
I don't have any idea what direction this will take. I was just describing the discussion I saw of the tension between keeping the model "pure", and trying to fund endless free requests.

Having an ad layer on top of the "objective" chatbot is easy, same way they handle tracking your token use and kicking you to a lower model (you can ask an LLM about it and it won't have visibility). But once you let people pay to influence the actual answers, that would seem like a dealbreaker for most.
 
JohnnyG said:
I don't have any idea what direction this will take. I was just describing the discussion I saw of the tension between keeping the model "pure", and trying to fund endless free requests.

Having an ad layer on top of the "objective" chatbot is easy, same way they handle tracking your token use and kicking you to a lower model (you can ask an LLM about it and it won't have visibility). But once you let people pay to influence the actual answers, that would seem like a dealbreaker for most.
Click to expand...
Wasn't trying to imply that *you* directly were assuming that it would pan out that way. For the record, I think they will do it just that way at first, and maybe even for quite some time. IIRC, once people started noticing that Google was actively manipulating search results, they lied about it at first. It actually took a few years until there was actual evidence, all the while you had people talking about how they had to attach reddit to actually get anything relevant. You'd think that actively manipulating results would be a dealbreaker, but there's still loads of people using Google.
 
You must log in or register to reply here.
Back
Top Bottom