OpenAI to recast hit chatbot as a route to higher-margin products before a potential IPO.
See full article...
See full article...
“Chat is dead”: OpenAI preps overhaul of ChatGPT
- Thread starter JournalBot
- Start date
Yep, keep piling more token use into the product. It sounds better but will end up making it prohibitively expensive when finally right priced.
Let the Tokenization Recession begin.
Let the Tokenization Recession begin.
Upvote
151
(155
/
-4)
Yup, come into the market, ruin normal web searches to the point that if I want any good results I have to use a chat bot, rug pull the chatbot and start charging out the nose.
Hyperscalers at their best.
Upvote
196
(200
/
-4)
Upvote
128
(131
/
-3)
I'm sure a billion users all want to code an app instead of ask for recipe ideas or all the other things normal people use ChatGPT for.
This is not gonna go well for them. Agent use might be more profitable but the market is smaller.
This is not gonna go well for them. Agent use might be more profitable but the market is smaller.
Upvote
117
(121
/
-4)
Sometimes we forget that search engines and what not are not really free.
Every time one search is done, data is collected and sold for different purpose.
I'm starting to smell a possible enshittification (is even possible) of basic services, so that you will have eventually to upgrade to paid ones, for anything.
They are spending money that they do not actually posses, based on projected income that it's very difficult to actually believe will be remotely close to the actual numbers.
Every time one search is done, data is collected and sold for different purpose.
I'm starting to smell a possible enshittification (is even possible) of basic services, so that you will have eventually to upgrade to paid ones, for anything.
They are spending money that they do not actually posses, based on projected income that it's very difficult to actually believe will be remotely close to the actual numbers.
Upvote
48
(50
/
-2)
Well, you see, Dario said in an interview that a theoretical version of Anthropic might be profitable if they just left all of their expenses off the books, so it must be true.
Edit: spelling
Last edited:
Upvote
51
(53
/
-2)
MilanKraft
Ars Tribunus Angusticlavius
"Chat is Dead..."
(But OpenAI still needs it / will still use it to draw in free users, in order to give them a taste of the related products they wish to sell, so)
..."Long live Chat" ... or something.
Also:
"It will transcend the actual surface . . . what we’re building towards is where you have your own personal agent that is capable of helping you… across everything in your life...."
The fucking EGO on these guys. They really think they're inventing God, don't they? Can't even speak plain English anymore. "AI is so... ABOVE, bro... so transcendent. You just can't see what we here at OpenAI see." This is the culture you get when it's created by a manipulative asshole like Altman.
What a hand-wavy crock of shit. Loosely translated: they have no idea what this amazing "final app / service" they're talking about will look like. Or which things it will do better than others, but they do know they can make it easy for people to assign a bunch of agents. So they make an all-encompassing, dramatic-sounding statement, when all they're really saying is, you'll have some sort of agent manager app to do a bunch of random shit for you.
(Whether each thing is a good idea or not they wouldn't know, as that requires actual hard-core testing for a significant period of time, and they don't give enough fucks to do that).
(But OpenAI still needs it / will still use it to draw in free users, in order to give them a taste of the related products they wish to sell, so)
..."Long live Chat" ... or something.
Also:
"It will transcend the actual surface . . . what we’re building towards is where you have your own personal agent that is capable of helping you… across everything in your life...."
The fucking EGO on these guys. They really think they're inventing God, don't they? Can't even speak plain English anymore. "AI is so... ABOVE, bro... so transcendent. You just can't see what we here at OpenAI see." This is the culture you get when it's created by a manipulative asshole like Altman.
What a hand-wavy crock of shit. Loosely translated: they have no idea what this amazing "final app / service" they're talking about will look like. Or which things it will do better than others, but they do know they can make it easy for people to assign a bunch of agents. So they make an all-encompassing, dramatic-sounding statement, when all they're really saying is, you'll have some sort of agent manager app to do a bunch of random shit for you.
(Whether each thing is a good idea or not they wouldn't know, as that requires actual hard-core testing for a significant period of time, and they don't give enough fucks to do that).
Last edited:
Upvote
108
(110
/
-2)
Translation: there will always be many distinct brands because AGI as we describe it is never happening, although we will use the label "AGI" for ($MajorVersion).0 where ($MajorVersion) is the release just before our IPO.
Upvote
27
(28
/
-1)
You mean it wasn't a big money maker to develop the go-to tool for college students to cheat on their essays, lawyers to fake their briefs, unscrupulous scientists to fabricate data/manuscripts, or others to flood the Internet with convincing-sounding nonsense? I can't believe it! /s
Last edited:
Upvote
66
(68
/
-2)
“Approximately a year ago, OpenAI’s strategy was swing for the fences, whereas Anthropic’s strategy is make money first,” said Jenny Xiao, partner at Leonis Capital and former researcher at OpenAI.”
Uh-huh, sounds reasonable.
“When we have [artificial general intelligence], I don’t think there will be a large number of distinct brands,” said Alex Embiricos, OpenAI’s head of enterprise product. “Probably there will be a single entity that I can talk to that can do whatever I need.”
So, how are those fences again?
Also I want to know if your AGI will finally be able to tell me how much milk is in my fridge because I feel that’s a serious gap in my life that technology has been unable to help me with despite all the grandiose promises.
Fuck me but I feel slightly stupider for having read that vacuous fluff piece masquerading as a serious article.
Uh-huh, sounds reasonable.
“When we have [artificial general intelligence], I don’t think there will be a large number of distinct brands,” said Alex Embiricos, OpenAI’s head of enterprise product. “Probably there will be a single entity that I can talk to that can do whatever I need.”
So, how are those fences again?
Also I want to know if your AGI will finally be able to tell me how much milk is in my fridge because I feel that’s a serious gap in my life that technology has been unable to help me with despite all the grandiose promises.
Fuck me but I feel slightly stupider for having read that vacuous fluff piece masquerading as a serious article.
Last edited:
Upvote
67
(68
/
-1)
Well, see, a less negative margin is technically a "higher" margin than an even more negative margin, if you plot on a graph.
Upvote
33
(34
/
-1)
I am honestly a bit of a sucker for AI stories and this one has got to be the stupidest thing I've read in a year.
Upvote
44
(45
/
-1)
900% of Zero silly.
But in all seriousness they need to look like they have some kind of path to a profitable future, because right now everyone is figuring out that AI is a 20-30 year proposition to make sense, and we're in year 5. The over leveraged investors aren't going to give them an Amazon sized runway to get their shit sorted. So they need to create the perception of value to hype the stock and rug pull before crashing the market and fleeing to somewhere with no extradition treaty.
The puffed up suits will talk about margin and profits as if they aren't fanciful future state items that will never happen before the bottom drops out.
So they will try pushing Agentic models and the like onto the few places that might have the money to spend hoping they can create the illusion of value for another quarter. Because even if you got every one on the planet paying $10 bucks a month that still wouldn't offset the trillions in value they claim to be worth. They need companies paying millions a month every month and more besides endlessly forever.
Upvote
27
(27
/
0)
The company I work for is already moving on from OpenAI products. Codex burned us with turning on 'fast' mode by default in a recent update, which has 2-2.5x token usage. With that, we coalesced around Claude Code.
Upvote
8
(8
/
0)
So I'm seeing this as OpenAI finally recognizing that LLMs have hit their technological limit and the only way they have to survive is to decorate the corpse of ChatGPT with enough baubles to attract attention.
Which probably means every AI vendor is in the same boat, they can't compete on LLM, they have to try and compete on shiny. And that is not ending well for any of them.
Upvote
39
(40
/
-1)
train_wreck
Ars Scholae Palatinae
This “article” is just a press release from OpenAI…. kinda gross to see here IMO
Upvote
39
(44
/
-5)
FT out here doing more to ruin journalism's credibility than trump could ever dream of. This is embarrassing, Cristina.
I think I'm going to train an AI solely on content intentionally stolen from FT, then use it to generate worthless emails that I'll subscribe FT "writers" to. Slop for the slop gods.
I think I'm going to train an AI solely on content intentionally stolen from FT, then use it to generate worthless emails that I'll subscribe FT "writers" to. Slop for the slop gods.
Last edited:
Upvote
25
(31
/
-6)
Try and remember that OpenAI's definition of "AGI" is not "acts like a rational entity," but "Makes $100,000,000,000."
Upvote
29
(29
/
0)
Re: milk
I had this idea, seriously actually, when I was about nine years old. Then, within a day or two, my nine year old self realized that if you changed a jug for a carton, or a different brand with thicker plastic, or put something that wasn't milk in the non-milk spot, or put the milk in the wrong spot, the whole thing fell apart.
Later - this wasn't the insight of a nine year old me - I knew you'd need scales, sensors, not just cameras but vision systems, at least thousands and maybe tens of thousands of dollars of stuff to be able to tell you that you're getting low on milk. Commoditization of much of that, including AI if you're okay with a few misses (vision is hard). Maybe get it down to a four or five thousand dollar refrigerator.
Or, you could pick up the carton.
Upvote
32
(32
/
0)
This is the techbro grifter cycle we see again and again.
Claim something (e.g LLMs) are the future of everything, milk every penny, and when the cracks and lies start to show suddenly it's dead, it's old news, the NEXT thing is actually where it's at.
It's always impossibly tasty jam tomorrow. It's Full Self Driving! No wait that's dead, it's Robotaxis! No wait, that's dead, it's personal Robots!
The speed at which their supporters are able to turn 180 from "this is the future" to "that was never the future, this other thing is" is astonishing, never stopping to question that the jam never arrives.
"Oceania had always been at war with Eastasia..."
Claim something (e.g LLMs) are the future of everything, milk every penny, and when the cracks and lies start to show suddenly it's dead, it's old news, the NEXT thing is actually where it's at.
It's always impossibly tasty jam tomorrow. It's Full Self Driving! No wait that's dead, it's Robotaxis! No wait, that's dead, it's personal Robots!
The speed at which their supporters are able to turn 180 from "this is the future" to "that was never the future, this other thing is" is astonishing, never stopping to question that the jam never arrives.
"Oceania had always been at war with Eastasia..."
Upvote
46
(47
/
-1)
Are chatbots going to enter into push advertising, that is getting into everything these days? /S
Upvote
4
(4
/
0)
Nope
https://isaiprofitable.com/
Edit: LLMs as a thing aren't going away, but these big companies are just scams all the way down.
Everyone who seems to be seriously questioning what they're doing and not just repeating their talking points seems to be finding that none of this shit adds up.
https://www.wheresyoured.at/ai-doesnt-have-roi/
https://justdario.com/2026/05/nvidia-is-now-very-close-to-hitting-the-circular-financing-brick-wall/
Last edited:
Upvote
22
(23
/
-1)
This really hit me over the weekend walking into a Lowe's. The big ridiculous riding mowers had tags on them, with "special offers" if you signed up. For a subscription.* To Lowe's. To buy stuff.
What?
*Pro accounts make more sense and come with management tools to track purchases, bulk/commercial discounts, etc.
Upvote
5
(5
/
0)
Their "product" is as transcendent as an outhouse, but having both their sense of smell and taste cauterized, they cannot tell. Their strategy was to "swing for the fences" but they fell so far short of AGI that they decided to tear down the fence, move a lot closer, and build a new fence only an inch tall. After all, no mere human could ever achieve booking travel or organizing a calendar, right? /S
Upvote
13
(14
/
-1)
Oh gods, I’d forgotten about that load of happy horseshit. From the article, it looks like Mr Embiricos has forgotten too, or at least he appears to have given that he’s jabbering on about entities that can do whatever he wants.
I will give him some extremely grudging credit for not following up with “and it’ll be mine, all mine, mwahahahaha!”
Upvote
9
(10
/
-1)
And never try to graph it to scale, your computer screen does not extend that far underground.
Upvote
5
(6
/
-1)
And so the goalposts move yet again. It seems 3Q and 4Q will be when suits industry-wide get the memo that tokens will ultimately cost more than the thousands they laid off. All this jockeying is madness.
Upvote
10
(11
/
-1)
Same old shit. 25 years ago people talked about the Segway and how cities would be redesigned for it.
Upvote
20
(21
/
-1)
So, who will take the time to connect ChatGPT or whatever it will be that's not chat but agents (nice thing about agents is that you can hire J. Edgar Boozer to fire them) with claude's agents and let them melt each other's data centers.
Upvote
-1
(0
/
-1)
"Actually Generates Income" is the AGI they've all been looking for.
It's all starting to make sense now.
Upvote
34
(34
/
0)