Hayao Miyasaki is the co-founder of Studio Ghibli, a Japanese animation studio known worldwide for their stunning, emotional, beautiful stories and movies. At the core of Studio Ghibli’s work is a deep engagement with questions of humanity. About what it means to be a human, about how to care for one another and the world […]
There is another aspect of this also. I could generate Ghibli style images a few years ago using better image generation models like stable diffusion or Midjourney. OpenAI is so lagging behind in terms of image generation it is comical at this point. But they get all the media coverage for these things as if they are inventing something out of thin air.
Most governments ignored the IP issues when other models were already doing these violations. Professionals are not using OpenAI. OpenAI only makes it so that these products reach big audiences. Then they become extremely accessible with the downside being that they are dumbed down. Thus, losing a lot of functionality.
This is what billionaires and major corporations are doing now and have been doing for a long time. Do you remember Titan sinking? What was so incredible is that the founder and CEO of Oceangate was acting like A: No one has ever gone to the Titanic before, and B: submarine travel is somehow a brand new thing that was just being invented by HIM.
This was utter bullshit on so many levels. James Cameron even spoke about how horrendous his assessment of the situation was, saying that the Titanic site is actually one of the riskier shipwrecks to go down to, which is why it needs to be approached with caution (which Oceangate did not care about), and that submarine travel is a very mature science and what the idiot CEO was doing wasn’t simply a bad idea in general, but he believed he could violate the laws of physics.
You can break the laws and rules of society, but you cannot break the laws of physics. If you jump off the top of a skyscraper, no amount of arm flapping will make you fly.
OpenAI is so lagging behind in terms of image generation it is comical at this point.
They dropped a new image model last week using 4o to contextualize the request, it’s very very good. However it’s for paid subscribers only right now I believe.
However as you mentioned Stable diffusion and mid journey probably still have more customizability.
There is another aspect of this also. I could generate Ghibli style images a few years ago using better image generation models like stable diffusion or Midjourney. OpenAI is so lagging behind in terms of image generation it is comical at this point. But they get all the media coverage for these things as if they are inventing something out of thin air.
Most governments ignored the IP issues when other models were already doing these violations. Professionals are not using OpenAI. OpenAI only makes it so that these products reach big audiences. Then they become extremely accessible with the downside being that they are dumbed down. Thus, losing a lot of functionality.
This is what billionaires and major corporations are doing now and have been doing for a long time. Do you remember Titan sinking? What was so incredible is that the founder and CEO of Oceangate was acting like A: No one has ever gone to the Titanic before, and B: submarine travel is somehow a brand new thing that was just being invented by HIM.
This was utter bullshit on so many levels. James Cameron even spoke about how horrendous his assessment of the situation was, saying that the Titanic site is actually one of the riskier shipwrecks to go down to, which is why it needs to be approached with caution (which Oceangate did not care about), and that submarine travel is a very mature science and what the idiot CEO was doing wasn’t simply a bad idea in general, but he believed he could violate the laws of physics.
You can break the laws and rules of society, but you cannot break the laws of physics. If you jump off the top of a skyscraper, no amount of arm flapping will make you fly.
They dropped a new image model last week using 4o to contextualize the request, it’s very very good. However it’s for paid subscribers only right now I believe.
However as you mentioned Stable diffusion and mid journey probably still have more customizability.