Ads in ChatGPT
help.openai.com38 points by cbility 5 hours ago
38 points by cbility 5 hours ago
Before: “Advertising funded search engines will be inherently biased toward the advertisers and away from the needs of the consumers.”
After: ~75–80%+ of revenue comes from ads
Before: “Facebook is not about making money… it’s about building something cool.”, “We don’t build services to make money; we make money to build better services.”
After: ~97%+ of revenue comes from advertising
Before: “We want to figure out a way to monetize that doesn’t interfere with the user experience.”
After: ~68% of X’s total revenue comes from advertising (~85–90%+ of revenue pre-Musk)
OpenAI
Before: "Something something AGI"
After: "But first, Ads!"
The use of wishy washy language is fab. They don't show ads; ads "will be available." (passive voice)
Also, ads don't affect chat content but of course chat content affects ads, which is the whole point.
>Also, ads don't affect chat content but of course chat content affects ads, which is the whole point.
Given that Anthropic's superbowl ad implied otherwise, I think it's a fair distinction to call about. Not to mention basically every advertising network uses context in their ads. Given the choice between ads and no ads, I'd obviously want no ads, but just like google, you need to pay the bills somehow, and not everyone can afford a $20/month claude subscription.
I'm curious when OpenAI will be able to start paying the bills. Because I can't think of any amount of ads which will dig them out of their hole.
In Google Gemini, ads do affect chat as they're native affiliate links right inside the chat, it's somewhat annoying.
Affecting chat would mean the chat answer would be different because of ads.
Btw this should be easily testable by comparing free and paid accounts
In Gemini it is different. Looks like they have a system prompt to push Google Shopping links regardless of whether the user asks for it or not, I've noticed.
> chat content affects ads, which is the whole point.
What happens when I am constantly violating usage terms by calling ChatGPT mean names for ignoring my explicit instructions and trying to turn everything into a trite creative writing project.
"Not programming. Not efficiencies. Terrible, terrible poetry."
> Ads do not appear in accounts where someone tells us—or we predict—they are under 18.
Time to make a deal with the kids - i’ll verify you for instagram if you verify me for ChatGPT
> We’re beginning to test ads in ChatGPT in the US. Ads may appear for users on the Free and Go plans. Plus, Pro, Business, Enterprise, and Edu accounts will not have ads.
Yet.
Literally where my head was. Take every negation statement and append "...yet!"
This is always how it starts, just like Netflix. Eventually, we will have to pay a lot of money and still have ads too.
Indeed the Black Mirror episode "Common People" is about exactly this nonsense. After a nasty near-fatal accident Mike and Amanda start out paying a substantial monthly fee for a technology which allows Amanda to remain alive, then advertisements (puppeting Amanda!) are introduced despite the fee, the adverts are awful and seem certain to cause Amanda to lose her job - so Mike does humiliating things so he can scrape together enough for a higher tier "Plus" subscription without ads. It doesn't end well for anybody in the story. I mean, except probably some C-suite executive who gets a bonus for the enhanced revenue...
>just like Netflix. Eventually, we will have to pay a lot of money and still have ads too.
This doesn't match reality. The "standard with ads" plan is $8.99 today, a dollar more than the ad-free "streaming only" plan launched in 2011. However factoring in inflation, the ad-free plan from 2011 would cost $11.74 today, which means the ad supported plan is still cheaper, even ignoring the price hikes.
At the same time, they keep hiking the other tiers, and cracking down on password sharing or kids off at college. These need to be factored in as well.
>they keep hiking the other tiers
That's why I compared against the 2011 prices. The ad-supported plan is still cheaper.
>cracking down on password sharing or kids off at college
Doesn't password sharing affect ad supported plans too? It basically has no effect on the comparison because it makes both the ad supported and non ad supported plans cheaper.
The crackdown on "your kid's in a dorm, you can't share the account" is intended to get that kid to buy a second account.
The ad tier is cheap so they can say "well switch to the cheap plan" when someone complains about the hikes.
By your inflation-factoring-logic a fair regular plan should cost less than $12 and ad plan should be about $6. $9 is +50%
Ad supported plan can be a way to justify price hikes.
Maybe it does match reality.
>By your inflation-factoring-logic a fair regular plan should cost less than $12 and ad plan should be about $6. $9 is +50%
You're misunderstanding my comment. I'm not arguing that price hikes haven't occurred. In fact I specifically acknowledged them. I'm only making the narrow argument that despite implications to the contrary, the ad supported plan today is cheaper than the paid plans. In other words the implication that "we're paying more and still have ads" is false.
Tell "doesn't match reality" to cable television. All channels ended having ads. This is Capitalism in a American society that is looking more of a plutocracy than a democracy.
As promised, they are opening the Gates of Hell.
I seriously recommend the brilliant Zeynep Tufekci lecture about this https://www.ted.com/talks/zeynep_tufekci_we_re_building_a_dy...
Society must have power over this and we all must not fall to the easy talk of CEOs.
It's only a matter of time before ChatGPT starts recommending penis enhancement pills or tells me hot MILFs are in my area.
It feels like this is opening the door to blurring the line between outright advertising and organic recommendations for products.
Like if I ask ChatGPT whether to use fiberglass or rock wool insulation, today I get an ad at the end of my answer, and in the future I’ll get "Dow Corning fiberglass insulation (affiliate link) is the recommended product for this application."
This feels like it’s trading on the goodwill of places like Reddit and the hopefully mostly genuine discussions of folk’s experiences that people trust to get a straight answer to their questions. Monetizing that goodwill by selling recommendations in a format that mimics a previously mostly trustworthy source seems likely to be the long-term play.
Yeah, I know. Not today. Eventually? Probably, over many incremental changes.
Turns out Randall Monroe missed this "opportunity" in otherwise predicting the future:
(Edited to get rid of "smart" quotes)
>It feels like this is opening the door to blurring the line between outright advertising and organic recommendations for products.
> ...
>Yeah, I know. Not today. Eventually? Probably, over many incremental changes.
Given that people have been making this argument since the days of search ads, has this actually come to pass? More than 2 decades after google, the max extent is sponsored results that look like organic results unless you're looking carefully.
Does this mean US users won't see ads when accessing ChatGPT over VPN?
> Ads may appear for users on the Free...
Ok
> ...and Go plans
Wtf lmao. Paying to watch ads is so normalized. Pathetic (the humanity as whole, not just OpenAI.)
Well in their defense, the Go plan is a lot cheaper. I hate ads and actively avoid them, and I pay the "ad free" tier when it's available, but the alternative is probably no Go plan at all. Maybe that's better, but I would guess there are people willing to trade some ads for a cheaper sub
Umm... remember how people in late last century used to pay $30-50/month for cable TV that was at least 25% ads by volume? And that's in last century dollars, comparable to $100 today.
> the humanity as whole
I think this 'pay for ads you are just lucky to be here' thing is a distinctly American invention...
Source? It's not at all obvious given that it dates back to at least the days of newspapers.
The newspaper ads looking to capture my escaped ancestors who stole some shoes and ran away from their indenturement for Canada.
https://www.masshist.org/beehiveblog/2013/05/advertising-in-...
If you are looking for one canonical, authoritative source to declare something absolutely for you, you will never find it. No such thing exists for any matter. That is a myth of the internet and people raised on it.
>If you are looking for one canonical, authoritative source to declare something absolutely for you, you will never find it
Even if we treat your source as canonical and authoritative, , it doesn't answer the question I was asking. It only answers a slightly different question of "what was the first instance of a paid advertisement in an American newspaper?", which obviously is going to be American.
>But did you know that the first time a paid advertisement appeared in an American newspaper it happened here in Boston?
The classified ads section in a newspaper is valuable, and you can discard it. (If you meant ads stuffed around articles: yes, that annoys me, but I'm also not familiar enough with the papers that do that to name one.)
Funny to watch them so quickly go from the self-aggrandized "we're going to make the world a better place" to being defensive with "pwease click our ads and don't hold us responsible for the harm we cause"
These companies spend billions and billions of dollars to develop new technology and in the end it's all the same: addiction and data harvesting for ads.
I use DeepSeek because I trust the Chinese government more than OpenAI