Does Apple start from scratch, too? OpenAI without MS doesn’t exist.
I think everyone thinks this is too easy.
Why is Google struggling when it has everything? It has the strongest supercomputing infrastructure. The infrastructure for their AI services and the cloud that lets them lend and manage such infrastructure to customers are completely different areas.
Investment in infrastructure for Generative AI is also time consuming. Even if a chip is made on its own, there is no answer if it receives the latest Nvidia chip and does not respond. AWS, which used to shout that it was making its own chip, hurriedly sends a message of reconciliation with Nvidia to the market, and sends a message that it will buy the most infrastructure and introduce and provide services as soon as possible.
It will take a lot of its own chip environment, but securing Nvidia chips is a priority right now. Everyone is not accepting this. It’s strange. Oracle worked with Nvidia to build GPU infrastructure first, but it’s fun these days. It’s nonsense to say that Oracle is dead. It’s reviving splendidly.
Due to Nvidia GPU Supercom in Oracle Cloud Infrastructure, there are rumors that starting with startups trying to learn Generative AI, it seems like a $100 deal has been made regarding Elon Must X AI learning that they were originally close to.
GPU infrastructure is the most basic and important. It won’t be too late to say the same thing as it will change in 10 years. Should I say that Samsung Electronics, which suffered a crushing defeat in the AI semiconductor market, is a scammer that it will come up soon? (I will respond at some point, but it will not be short-term. I’m not saying that I’m not looking for a new breakthrough right now. ^.^)
Why did TensorFlow lose to the Pytorch camp created by Meta? They made it to use their own TPU, so of course, they’re less flexible when using other chips. They made it just for themselves.
Microsoft Azure is in charge of all the infrastructure of OpenAI’s innovation. Their cooperation with Apple means that Apple has also joined hands with Microsoft.
Google has 1 billion search users. To provide all of the Generative AI to these targets, an unimaginable AI infrastructure is required. In comparison, Microsoft and OpenAI have been providing services to 100 million people for more than a year. It was not an exaggeration to say that 100 million people is not a small number, but perhaps it has the experience of operating infrastructure that provides the most powerful service. At the same time, you can gradually expand your infrastructure. While designing new chips.
The problem is cooperation with Apple.
The iPhone was first released in 2007, and its users continue to grow every year. There are many market institutions that believe it has reached or reached approximately 2 billion by 2023.
How much infrastructure does it need to be to provide Generative AI services to those users? Even if Apple pays Nvidia unimaginable money to supply it first, it cannot immediately introduce that infrastructure and serve those people.
It is a company that gives 20 billion dollars to Google through its exclusive use of iPhone search. If it cooperated with OpenAI, of course, it cannot be explained without considering Microsoft’s infrastructure. However, it is not easy for Microsoft to say that it will give it free of charge. This is because the U.S. Fair Trade Regulatory Agency is currently filing an anti-trust lawsuit against Google search, and Microsoft is vilifying Google, saying that it is bad for Google. Seeing how he made a deal with Apple using OpenAI while acting like that, I think he has found a clever move.
How to avoid the blade of each country’s regulators heading to Microsoft by attracting Apple to its stake in OpenAI or an exclusive contract between Microsoft and OpenAI. I wonder what it is.
It seems difficult for Apple to release all related services for free. Google and Samsung Electronics also solved Generative AI service for free only for new products. That was also temporary.
Apple is already providing a fee in the service area, so it is not necessary to make a monthly fee for Generative AI. Microsoft is clearly positioned as B2B and is often cooperating with Apple. Open AI and Microsoft are said to have different minds, but I don’t think AI companies without large GPU infrastructure can escape cloud operators now.
OpenAI lost raw oil to the Nvidia team when it was nearing the end of the GPT-4o demonstration. I also thought that maybe the new Nvidia product that was released last time was provided first.
Meanwhile, the posting was posted on Microsoft Azure immediately after the service introduction.
Introducing GPT-4o: OpenAI’s new flagship multimodal model launches preview on Azure
By the way, Google had developer events and cloud events one month apart. I don’t know why they did it separately. Companies have to do a developer event and then a B2B event to better prepare, but this has changed the order. I think I’m losing my mind.
The developer I met a few days ago said Google seems to have a good direction, but something seems to be missing. He said the quality was not very good either. Maybe it’s because they’re geniuses, so they’re not used to kneeling down and serving others. A former Oracle came and told me to relax, so I’m making profits, but I still have a long way to go.
It is questionable whether Google will be able to shout as loudly as it is now when it is reeling from the anti-trust lawsuit in search. Given that OpenAI is doing something to create a search service, Google should also be on alert. Sam Altman seems to want to go down in history as the person who beat Google. Even if he uses his backseat to Microsoft, he will put his sword in the hands of Microsoft. Now, I think it is urgent to join hands with a huge customer named Apple to put Google in a corner.
a
I think everyone thinks this is too easy.
Why is Google struggling when it has everything? It has the strongest supercomputing infrastructure. The infrastructure for their AI services and the cloud that lets them lend and manage such infrastructure to customers are completely different areas.
Investment in infrastructure for Generative AI is also time consuming. Even if a chip is made on its own, there is no answer if it receives the latest Nvidia chip and does not respond. AWS, which used to shout that it was making its own chip, hurriedly sends a message of reconciliation with Nvidia to the market, and sends a message that it will buy the most infrastructure and introduce and provide services as soon as possible.
It will take a lot of its own chip environment, but securing Nvidia chips is a priority right now. Everyone is not accepting this. It’s strange. Oracle worked with Nvidia to build GPU infrastructure first, but it’s fun these days. It’s nonsense to say that Oracle is dead. It’s reviving splendidly.
Due to Nvidia GPU Supercom in Oracle Cloud Infrastructure, there are rumors that starting with startups trying to learn Generative AI, it seems like a $100 deal has been made regarding Elon Must X AI learning that they were originally close to.
GPU infrastructure is the most basic and important. It won’t be too late to say the same thing as it will change in 10 years. Should I say that Samsung Electronics, which suffered a crushing defeat in the AI semiconductor market, is a scammer that it will come up soon? (I will respond at some point, but it will not be short-term. I’m not saying that I’m not looking for a new breakthrough right now. ^.^)
Why did TensorFlow lose to the Pytorch camp created by Meta? They made it to use their own TPU, so of course, they’re less flexible when using other chips. They made it just for themselves.
Microsoft Azure is in charge of all the infrastructure of OpenAI’s innovation. Their cooperation with Apple means that Apple has also joined hands with Microsoft.
Google has 1 billion search users. To provide all of the Generative AI to these targets, an unimaginable AI infrastructure is required. In comparison, Microsoft and OpenAI have been providing services to 100 million people for more than a year. It was not an exaggeration to say that 100 million people is not a small number, but perhaps it has the experience of operating infrastructure that provides the most powerful service. At the same time, you can gradually expand your infrastructure. While designing new chips.
The problem is cooperation with Apple.
The iPhone was first released in 2007, and its users continue to grow every year. There are many market institutions that believe it has reached or reached approximately 2 billion by 2023.
How much infrastructure does it need to be to provide Generative AI services to those users? Even if Apple pays Nvidia unimaginable money to supply it first, it cannot immediately introduce that infrastructure and serve those people.
It is a company that gives 20 billion dollars to Google through its exclusive use of iPhone search. If it cooperated with OpenAI, of course, it cannot be explained without considering Microsoft’s infrastructure. However, it is not easy for Microsoft to say that it will give it free of charge. This is because the U.S. Fair Trade Regulatory Agency is currently filing an anti-trust lawsuit against Google search, and Microsoft is vilifying Google, saying that it is bad for Google. Seeing how he made a deal with Apple using OpenAI while acting like that, I think he has found a clever move.
How to avoid the blade of each country’s regulators heading to Microsoft by attracting Apple to its stake in OpenAI or an exclusive contract between Microsoft and OpenAI. I wonder what it is.
It seems difficult for Apple to release all related services for free. Google and Samsung Electronics also solved Generative AI service for free only for new products. That was also temporary.
Apple is already providing a fee in the service area, so it is not necessary to make a monthly fee for Generative AI. Microsoft is clearly positioned as B2B and is often cooperating with Apple. Open AI and Microsoft are said to have different minds, but I don’t think AI companies without large GPU infrastructure can escape cloud operators now.
OpenAI lost raw oil to the Nvidia team when it was nearing the end of the GPT-4o demonstration. I also thought that maybe the new Nvidia product that was released last time was provided first.
Meanwhile, the posting was posted on Microsoft Azure immediately after the service introduction.
Introducing GPT-4o: OpenAI’s new flagship multimodal model launches preview on Azure
By the way, Google had developer events and cloud events one month apart. I don’t know why they did it separately. Companies have to do a developer event and then a B2B event to better prepare, but this has changed the order. I think I’m losing my mind.
The developer I met a few days ago said Google seems to have a good direction, but something seems to be missing. He said the quality was not very good either. Maybe it’s because they’re geniuses, so they’re not used to kneeling down and serving others. A former Oracle came and told me to relax, so I’m making profits, but I still have a long way to go.
It is questionable whether Google will be able to shout as loudly as it is now when it is reeling from the anti-trust lawsuit in search. Given that OpenAI is doing something to create a search service, Google should also be on alert. Sam Altman seems to want to go down in history as the person who beat Google. Even if he uses his backseat to Microsoft, he will put his sword in the hands of Microsoft. Now, I think it is urgent to join hands with a huge customer named Apple to put Google in a corner.