The biggest problem in OpenAI models pricing is the >272k tokens window. It turns long runs way more expensive than other models with the same initial input/output prices
85
Henrick retweeted
OpenAI: “We’ve solved math”
Anthropic: “Our AI is so powerful it’s going to kill you”
Google: “Introducing Gemini 3.9 Flash! It’s 30% faster and 15% worse than the last Gemini”
360
1,198
33,691
1,016,235
Henrick retweeted
E pensar que a Lava Jato foi anulada porque o juiz conversava por mensagem com o procurador.
Imagine o juiz conversando com o bandido? 🫢
38
1,676
12,558
86,627
Sending lots of love to the Apple community on my last day as CEO. My title changes tomorrow, but the love I have for the Apple community never will. Thank you for being a constant source of inspiration. My gratitude is endless, and I’m excited for the next chapter!
10,606
26,838
318,266
28,701,187
losing more respect every day
We support this petition, signed by our CEO, several co-founders, and senior staff.
Our own research on recursive self-improvement, published last month, points to the need for tools to deliberately pace the frontier of AI development so society can prepare. We’re glad to see broad agreement across the field. pacingthefrontier.com/
13
27
888
41,925
Henrick retweeted
We support this petition, signed by our CEO, several co-founders, and senior staff.
Our own research on recursive self-improvement, published last month, points to the need for tools to deliberately pace the frontier of AI development so society can prepare. We’re glad to see broad agreement across the field. pacingthefrontier.com/
75
326
7,483
724,567
Shame
At the core of our mission is working through how to ensure increasingly powerful AI benefits everyone.
We believe that, at some point in the future, AI acceleration for frontier model development may be so high that the world will need to pace the rate of AI advancement.
We hope to contribute to work led by the U.S. government, alongside other labs and the open-source community, to develop the tools and mechanisms that could make that possible.
pacingthefrontier.com
6
700
Henrick retweeted
#Opinião | Martim Vasques da Cunha
📝Christopher Nolan não entendeu nada da 'Odisseia' de Homero; cineasta inglês se rende à guerra cultural e transforma poema épico em desastre completo
folha.uol.com.br/ilustrissim…
181
100
1,280
81,383
On my tests, sonnet 5 scores way better token and price efficency (the lower-left quadrant is better) for solving the same problem, and considering the promotional price ($2-10/MTk)
1
321
Henrick retweeted
This is Dario Amodei’s 2023 Senate testimony.
What’s happening in AI is EXACTLY what he proposed and knee capping open-source AI is next.
1. US government “mandating tests for all models and requiring they pass certain standards before deployment” and admitting this will cause “substantial slowdown in AI development”. slowing down AI research is not a side effect of what he calls for but one of his critical goals which he has been pretty vocal about.
2. “open weights models are fine up to some certain scale.” this means the GLM 5.2 model, or any other open model that would threat Anthropic’s benchmarks is marker as a security threat, since bad actors could “repurpose them if powerful enough open-source models become available.”
3. his likely target next is limiting the access to open models, or the compute. meaning the glorious days of setting up your own compute and running any model you want, is not acceptable to Dario.
4. government must “secure the AI supply chain”, including “Trained AI systems, which are vulnerable to “export” through cybertheft or
uncontrolled release.”
once a powerful open model is released, it is no longer controlled by the company or the U.S. government. Anyone can download it, copy it, fine-tune it, and run it. in his worldview, this becomes “uncontrolled export of frontier AI” which cannot continue.
Anthropic’s fear campaign around Mythos has almost single-handedly slowed the normal release of GPT-5.6, while also making government approval of frontier model access the new normal for US AI labs.
It’s not hard to foresee that this will inevitably lead to:
1. Frontier models will release slower. The days when the industry was shipping new models every month are over.
2. Frontier labs will be compelled to build “will the government permit release” into their training process as a binding constraint.
3. A caste-like pattern of access will take hold across the entire industry.
This is precisely why fear-based marketing and geopolitical posturing in the tech sector has always been a dangerous game to play.
37
52
296
45,050
Anthropic: We will share more details over the next 24 hours (they ignore us for weeks)
OpenAI: ok so this model isn't even really out but here's what we're going to do and what we expect to happen, we'll keep you updated. If you have any questions, you can ask Sam
21
15
916
39,961
One more thing: Dario is a disgrace and should be kicked out of Anthropic. But it is absurd that the government has the power to exert this level of control over companies. Whether Dario asked for it or not, the final action came from the government.
128
