> Maybe even more importantly: Anthropic wants to control what people do with AI—they block companies they don't like from using their coding product (including us), they want to write the rules themselves for what people can and can't use AI for, and now they also want to tell other companies what their business models can be.
What you cannot do. You may not use our Services for any illegal, harmful, or abusive activity. For example, you may not:
1. Use our Services in a way that infringes, misappropriates or violates anyone’s rights.
2. Modify, copy, lease, sell or distribute any of our Services.
3. Attempt to or assist anyone to reverse engineer, decompile or discover the source code or underlying components of our Services, including our models, algorithms, or systems (except to the extent this restriction is prohibited by applicable law).
4. Automatically or programmatically extract data or Output (defined below).
5. Represent that Output was human-generated when it was not.
6. Interfere with or disrupt our Services, including circumvent any rate limits or restrictions or bypass any protective measures or safety mitigations we put on our Services.
7. Use Output to develop models that compete with OpenAI.
Learning: Marketing and Legal team are not in sync with each other.
Sam's goal is to create viral products. This has always been his goal. While his interest in and passion for AI are almost certainly genuine, he famously hedged his bets before committing to OpenAI up to the point of being ambiguously fired from his previous job. ChatGPT was one of the greatest viral products of all time. His subsequent attempts to create virality haven't been working in any real, durable way.
Enter Claude Code. It went viral, not in the awe-inspiring world-changing way that ChatGPT did, but in a way that allowed it to decisively overwhelm OpenAI's advantage in some very valuable niches. It wasn't born out of a search for a viral product, but rather out of an intellectual obsession with safe, steerable AI systems -- an obsession that could've just as easily ended up producing no products at all. It's the kind of obsession that can't be replicated if the only motive is virality.
So yeah, this is going to bother him a lot. He's defending ads while his competitor's product is having its "iPhone moment." You know this is part of the raw emotional driver because he ends his defense of ads by talking about Codex app downloads and saying he "believe[s] Codex is going to win," implying that significant doubt exists that he feels the need to correct.
Counterpoint being that Slack, for example, for all its faults, does not have ads in its chats.
If Anthropic is positioned as "thing for professionals to do professional work" then I think you just avoid this issue entirely. Fee for service. OpenAI trying to be the thing everyone is using won't work in that model, though.
Oh dear. The ad doesn’t directly reference openai at all and for those of us out of the loop on all of this we wouldn’t have read it as such. I had no idea openai were considering this.
I think Sam would have been far better off to let this lie. I’m not at all sure of the nature of the ads coming in chatgpt but by responding so personally and aggressively to what appears to be sillyness feels like they hit a bit close to home?
Yeah I just don't think the average person is going to even get what they are talking about and they may not really care about ads in exchange for free-to-use ChatGPT. Most ChatGPT users are just using it as a search and summary tool. And they were seeing ads before via Google so what difference does it make?
While the Anthropic ads might be a hit amongst nerds, it won't compel normies to use Claude over ChatGPT, and in fact, it may just be damaging to AI adoption overall as it presents bad taste examples of how corruptible it all is.
Great idea! A tweet providing thoughtful commentary on a competitors ads will surely set the record straight, people always respect CEOs that are willing to publicly talk about touchy topics. Would you like me to draft one for you?
Sam is an entrepreneur trying to play an industrialist game. He will lose. OpenAi will be sold to Microsoft on the profit side and control of the open source board will effectively be awarded to Elon. OpenAi had their chance to oust him, but his c-suit judo was too strong.
Of course the first version of the ad will be separated very clearly, just like how Google search started.
While I found the golden dating app the funniest anc creepiest, credit loan ad was the most viewed for a reason: it's the most profitable ad from the scenarios, and it will come in a few years.
I write all this as a mostly happy OpenAI subscriber (my only wish is still being able to have a sticky legacy o3 model as my default for my Pro subscription everywhere).
I don’t care for Sam Altman but so far I have been impressed with Codex and at least for our use right now its performing much better than Opus 4.5. Of course this will probably change again but the larger context window with Codex is god sent for us.
It feels like working with a professional. It just keeps churning until the work is done, and actually is pretty damn compact with token usage. Definitely lowest output tokens to value of the frontier models.
> Maybe even more importantly: Anthropic wants to control what people do with AI—they block companies they don't like from using their coding product (including us), they want to write the rules themselves for what people can and can't use AI for, and now they also want to tell other companies what their business models can be.
OpenAI's terms of service (https://openai.com/policies/row-terms-of-use/):
What you cannot do. You may not use our Services for any illegal, harmful, or abusive activity. For example, you may not:
1. Use our Services in a way that infringes, misappropriates or violates anyone’s rights.
2. Modify, copy, lease, sell or distribute any of our Services.
3. Attempt to or assist anyone to reverse engineer, decompile or discover the source code or underlying components of our Services, including our models, algorithms, or systems (except to the extent this restriction is prohibited by applicable law).
4. Automatically or programmatically extract data or Output (defined below).
5. Represent that Output was human-generated when it was not.
6. Interfere with or disrupt our Services, including circumvent any rate limits or restrictions or bypass any protective measures or safety mitigations we put on our Services.
7. Use Output to develop models that compete with OpenAI.
Learning: Marketing and Legal team are not in sync with each other.
My read:
Sam's goal is to create viral products. This has always been his goal. While his interest in and passion for AI are almost certainly genuine, he famously hedged his bets before committing to OpenAI up to the point of being ambiguously fired from his previous job. ChatGPT was one of the greatest viral products of all time. His subsequent attempts to create virality haven't been working in any real, durable way.
Enter Claude Code. It went viral, not in the awe-inspiring world-changing way that ChatGPT did, but in a way that allowed it to decisively overwhelm OpenAI's advantage in some very valuable niches. It wasn't born out of a search for a viral product, but rather out of an intellectual obsession with safe, steerable AI systems -- an obsession that could've just as easily ended up producing no products at all. It's the kind of obsession that can't be replicated if the only motive is virality.
So yeah, this is going to bother him a lot. He's defending ads while his competitor's product is having its "iPhone moment." You know this is part of the raw emotional driver because he ends his defense of ads by talking about Codex app downloads and saying he "believe[s] Codex is going to win," implying that significant doubt exists that he feels the need to correct.
Actual ad since it's not available in the tweet: https://www.youtube.com/watch?v=kQRu7DdTTVA
They're making a big promise here, that very few tech companies have been able to keep in the past.
Maybe there's a predictive market gamble starting about how long it will take Claude to follow suit if OpenAI starts making 9 figures in ad revenue.
Counterpoint being that Slack, for example, for all its faults, does not have ads in its chats.
If Anthropic is positioned as "thing for professionals to do professional work" then I think you just avoid this issue entirely. Fee for service. OpenAI trying to be the thing everyone is using won't work in that model, though.
Oh dear. The ad doesn’t directly reference openai at all and for those of us out of the loop on all of this we wouldn’t have read it as such. I had no idea openai were considering this.
I think Sam would have been far better off to let this lie. I’m not at all sure of the nature of the ads coming in chatgpt but by responding so personally and aggressively to what appears to be sillyness feels like they hit a bit close to home?
Yeah I just don't think the average person is going to even get what they are talking about and they may not really care about ads in exchange for free-to-use ChatGPT. Most ChatGPT users are just using it as a search and summary tool. And they were seeing ads before via Google so what difference does it make?
While the Anthropic ads might be a hit amongst nerds, it won't compel normies to use Claude over ChatGPT, and in fact, it may just be damaging to AI adoption overall as it presents bad taste examples of how corruptible it all is.
Yea, the response really sounds ultra defensive, almost like Anthropic nailed exactly what they're planning to do, and it's now Damage Control time...
Incredible! Reminds me of those fake GI Joe PSAs from 20 years ago.
An ad about not serving ads.
>One authoritarian company won't get us there on their own, to say nothing of the other obvious risks. It is a dark path.
It warms my heart to hear that Sam is against authoritarianism. Hopefully he doesn't hang out with anyone that supports that kind of thing.
https://www.wsj.com/tech/ai/the-real-story-behind-sam-altman...
I would have run this one past ChatGPT before hitting send.
Great idea! A tweet providing thoughtful commentary on a competitors ads will surely set the record straight, people always respect CEOs that are willing to publicly talk about touchy topics. Would you like me to draft one for you?
Deleting paragraphs 2 and 3 and tweaking paragraph 4 could have made this a great and informative reply by Altman.
That would favorably contrast ChatGPT's products, pricing, and market size with Claude's.
Instead, well, it's not a good look.
Sam is an entrepreneur trying to play an industrialist game. He will lose. OpenAi will be sold to Microsoft on the profit side and control of the open source board will effectively be awarded to Elon. OpenAi had their chance to oust him, but his c-suit judo was too strong.
I can’t put my finger on it but he’s just such an unlikeable figure
https://xcancel.com/sama/status/2019139174339928189
> We are building safe and beneficial AGI, but will also consider our mission fulfilled if our work aids others to achieve this outcome.
Maybe it's just me, but sometimes it feels like Sam does not actually want to aid others in achieving this outcome.
Of course the first version of the ad will be separated very clearly, just like how Google search started.
While I found the golden dating app the funniest anc creepiest, credit loan ad was the most viewed for a reason: it's the most profitable ad from the scenarios, and it will come in a few years.
I write all this as a mostly happy OpenAI subscriber (my only wish is still being able to have a sticky legacy o3 model as my default for my Pro subscription everywhere).
> But I wonder why Anthropic would go for something so clearly dishonest.
Mr. Dishonest accusing others of being dishonest.
> More importantly, we believe everyone deserves to use AI and are committed to free access, because we believe access creates agency.
I think you meant "psychological dependence", not agency.
Discussion (53 points, 77 comments) https://news.ycombinator.com/item?id=46894151
Defensive much, Sam?
This tweet must have made Anthropic's day.
The response is good if this means ads like that will never happen.
Google though... they make fake ad Emails so fake ad Replies seems exactly like their style.
While I love to see billionaires duking it out. I am rooting for Chinese open source models that I can run on my MacBook.
Now I want to find and watch the Anthropic advertisement.
I don’t care for Sam Altman but so far I have been impressed with Codex and at least for our use right now its performing much better than Opus 4.5. Of course this will probably change again but the larger context window with Codex is god sent for us.
It feels like working with a professional. It just keeps churning until the work is done, and actually is pretty damn compact with token usage. Definitely lowest output tokens to value of the frontier models.
Surely hit a nerve.