Support NeoGAF

[Draugoth]

OpenAI has confirmed ChatGPT will begin displaying advertisements, marking a significant pivot for the AI platform that once positioned itself as an alternative to ad-saturated search experiences. The move, announced through a company blog post, signals the need to offset mounting operational costs reportedly reaching around $5 billion annually, while managing a valuation approaching $300 billion following recent funding rounds.

The advertising rollout will predictably begin with ChatGPT's free tier users, though OpenAI has indicated paid subscribers may eventually see ads as well. The company stated it aims to make advertisements "relevant and useful," suggesting a targeted approach that mirrors established digital advertising models rather than introducing a new or novel monetisation framework.

The backlash

On Reddit, users are not happy with these ads that have popped up seemingly out of nowhere. One user said, "The ads thing is pretty b******t. I might just cancel my Pro subscription."

Another added, "I have been seeing ads as a pro user for at least 2 to 3 days now. (UK user, so it may be region-specific at the moment)."

One user chose to clarify that these "ads" have been visible for a while now, "ChatGPT is not showing ads. I imagine this refers to this instance in which Pro users received app connection suggestions, but that is not new. If you are a designer, you may receive a Canva app connection suggestion, sure. But what people understand as ads is not this, and this is not new this week. What has been new is the leak showing OpenAI was preparing the actual ad rollout, but we have not seen it in practice."

Article:

The era of ads in ChatGPT begins – users furious as even $200 a month Pro subscribers hit with app suggestions

Source: https://www.techradar.com/ai-platforms-assistants/chatgpt/the-era-of-ads-in-chatgpt-begins-users-furious-as-even-usd200-a-month-pro-subscribers-hit-with-app-suggestions

 

[Bojji]

 

[JeTmAn81]

They have to start making money now

 

[Paasei]

 

[BennyBlanco]

Beginning of the end hopefully. They desperately need to make money that isn't another corp handing them money.

 

[T8SC]

Looks like a targeted ad.

 

[DeepEnigma]

 

[winjer]

Seems of desperation, as the billions invested keep pilling up, with no return in sight.

 

[rm082e]

HSBC projects that even if OpenAI is making $200 billion by 2030, it would still need a staggering $207 billion in funding to stay afloat. As OpenAI scales, so does its costs. Its expensive frontier models like Sora 2 and GPT-5 cost millions of dollars a day to run, requiring enormous amounts of compute. These models are given away at cost in attempts to drive adoption, create lock-in, and essentially alter human behavior. You can think of it similarly to Spotify, which spent well over a decade completely unprofitable until it had basically re-written the way humans consume music and how labels distribute music. OpenAI is trying to achieve a similar result on a far, far larger, and potentially far more damaging scale.

Source

The math isn't mathing, and even if they stuff a shit ton of ads into their services, it's not going to be anywhere near enough.

 

[Jinzo Prime]

Just remember that the 3rd biggest company on Earth, Google, gets 70% of it's revenue from advertising (down from >90%). It's all about ads, that's where the money is at for these kind of companies.

 

[Mikado]

On the plus side, it will be even easier to catch people out when they just lazypaste ai slop into their essays and articles. since they'll probably also copy the ad text.

 

[near]

I haven't seen any ads yet, but the moment I do, I'll cancel my sub and try something else.

 

[Dacvak]

I haven't seen any ads yet, but the moment I do, I'll cancel my sub and try something else.

Same.

 

[Zathalus]

Desperation to make money.

 

[DKehoe]

Paying $200 a month and still getting ads is mental.

 

[Salz01]

I get the ads on free tier, but on paid fuck that…. I'll continue to build bots on an api till they figure out how to push ads through it. Then I'll go full open source.

 

[nbcjr]

to be honest, i understand they need to get money from somewhere, I do hope they succeed, otherwise only gemini will survive.

 

[AJUMP23]

Everything gets monetized.

 

[RiccochetJ]

Not shocked. They're burning through billions and they have to start recuperating these costs somehow. And based on news the other day agentic platform sales in the corporate world aren't materializing at the rate any of these AI vendors would like.

 

[Wilhelm_85]

I haven't seen any ads yet, but the moment I do, I'll cancel my sub and try something else.

Same. ChatGPT is great at a lot of things, but AI tools are already basically a commodity at this point for general use. I currently pay for Grok and ChatGPT (just the lowest levels), but I'll happily leave ChatGPT is they start serving me ads.

 

[DragoonKain]

The Black Mirror episode Common People might have been the most prescient of any Black Mirror episode. I know it doesn't specifically rate to ChatGPT or this form of AI, but in the greater picture you can understand what I mean. I appreciate that episode more and more everyday.

 

[6502]

How long until we get product placement in ai stories / text?

 

[Spyxos]

I was wondering why there haven't been any commercials for so long.

 

[Trilobit]

I can't wait for my house robot to restrain me when I'm trying to go to the toilet during an ad break.

 

[Big-G]

Gotta make that AI profitable for investors. All this money coming in, and there is more scrutiny to perform.

 

[SJRB]

This is a long time coming but serving ads to subscribers is diabolical.

 

[IDKFA]

Seems of desperation, as the billions invested keep pilling up, with no return in sight.

This.

Billions upon billions invested, but very little returns to show for it.

soon

 

[Forsete]

Haha, the bubble is starting to pop.

Enjoy the ride!

 

[JimmyRustler]

Haha, the bubble is starting to pop.

Enjoy the ride!

From what I've read the other day it's more like that ChatGPT is shedding customers to the competion.
I'm kinda getting a bit jaded with it lately as well, especially when you want it to help you solve a slightly more difficult problem and it keeps spewing you wrong answers. Like the other day I was trying to get it help me calibrate my new TV and just had to give up at some point. No matter how often I told it I don't have the menu it's describing, it kept telling me the same thing, just slightly different.

Was the first time I actually experianced the downside of AI. Unlike a human being it will never consult someone else if it doesn't know. It'll just keep repeating itself.

 

[Draugoth]

Haha, the bubble is starting to pop.

Enjoy the ride!

I give it 2-3 years still. But it will be catastrophic once it happens.

 

[MaestroMike]

 

[kpopSuperstar]

 

[Sand Dan Glokta]

Eh, don't care. Deepseek is superior to it in every regard anyhow.

 

[Hypereides]

Might be me, but it seems like a piece in the Jenga tower foundation is beginning to wobble...

Eh, don't care. Deepseek is superior to it in every regard anyhow.

Sure, if you prefer a censored pro-CPP slant in your suggestions.

For the record; all current AI/LLM models have some sort of slant to them.

 

[Sand Dan Glokta]

Might be me, but it seems like a piece in the Jenga tower foundation is beginning to wobble...


Sure, if you prefer a censored pro-CPP slant in your suggestions.

For the record; all current AI/LLM models have some sort of slant to them.

This sentence completely cancels out the one above lol. Exchanging one propaganda with another.

 

[Hypereides]

This sentence completely cancels out the one above lol. Exchanging one propaganda with another.

Not sure what you mean. In that post, I acknowledged they all carry some sort of bias/slant so that folks can be aware of about that and make an informed decision instead of trusting them all blindly.

I would've infinitely preferred one that was entirely neutral.

 

[Goalus]

Ramblings of a one-trick pony.

 

[Sand Dan Glokta]

Not sure what you mean. In that post, I acknowledged they all carry some sort of bias/slant so that folks can be aware of about that and make an informed decision instead of trusting them all blindly.

I would've infinitely preferred one that was entirely neutral.

Yes I agree, but you started with saying deepseek has biases for ccp, indicating somehow it makes it less viable, while chatgpt has opposite biases. Neither you nor anyone else brought it up till i mentioned deepseek, which is superior in my experience as i said. I just call it all built in propaganda.

 

[Salz01]

If they want to make money, release the NSFW mode….

 

[Hypereides]

Yes I agree, but you started with saying deepseek has biases for ccp, indicating somehow it makes it less viable, while chatgpt has opposite biases. Neither you nor anyone else brought it up till i mentioned deepseek, which is superior in my experience as i said. I just call it all built in propaganda.

Leaning on DeepSeek alone is unreliable and invalid, as far as I'm concerned.

The complete context, and my intended message, was that these things should all be approached with caution and can not be trusted on their own. Its better to cross-check information through more than one chatbot or compare. I'm drilling this home more and more since there's a real risk of people unintentionally believing or spreading disinformation and polarizing material if there's a lack of awareness.

 

[Sand Dan Glokta]

Leaning on DeepSeek alone is unreliable and invalid, as far as I'm concerned.

The complete context, and my intended message, was that these things should all be approached with caution and can not be trusted on their own. Its better to cross-check information through more than one chatbot or compare. I'm drilling this home more and more since there's a real risk of people unintentionally believing or spreading disinformation and polarizing material if there's a lack of awareness.

I'm not sure why the conversation led to disinformation or polarizing material when deepseek was mentioned...you don't even know for what/when/how much I use it...

I understand you core message, but it comes off as very patronizing, just saying.

 

[Hypereides]

I'm not sure why the conversation led to disinformation or polarizing material when deepseek was mentioned...you don't even know for what/when/how much I use it...

I understand you core message, but it comes off as very patronizing, just saying.

I dunno, dude. You threw the claim that my original post was self-contradictive from your view. I just expanded on my reasoning behind it.

But, I guess I'll just leave the convo here.

 

[ForAcademicPurposes]

 

[dsp]

 

[Coconutt]

I would immensely enjoy if forced ads truly became the downfall for Ai.

 

[simpatico]

As the world's only Co-pilot user, I'm not affected yet.

 

[Hot5pur]

People put up with ads on social media and search so I don't see how this will be any different.
For the average user AI is just fancy Google search.
For people who use it seriously at work for code or analysis, likely it will be a paid tier.
It's a great tool, it's powerful and a big time saver in very specific instances, but it's clear we're at the point where things are not changing much despite whatever gamed benchmarks will tell you.
This ain't a trillion dollar industry, but definitely something that could be profitable and sustainable for specific use cases.

 

[ReBurn]

As the world's only Co-pilot user, I'm not affected yet.

I use CoPilot in VSCode pretty much every day. It's a great assistance tool. I like being able to switch between OpenAi and Anthropic models pretty much on the fly.

 

[Sand Dan Glokta]

I use CoPilot in VSCode pretty much every day. It's a great assistance tool. I like being able to switch between OpenAi and Anthropic models pretty much on the fly.

Could you elaborate more on what you mean?

 

[AREYOUOKAY?]

Oh wait this thread is about getting ads... My mistake.