Article directory
1. ChatGPT research and development upsurge
Since ChatGPT sounded the clarion call of the artificial intelligence revolution again, the "Hundred Models War" has also begun on both sides of the Pacific Ocean.
The powerful generative conversational capabilities of OpenAI ChatGPT have sparked renewed interest and investment in AI in recent months. With the upsurge of ChatGPT-like research and development at home and abroad, conversational AI and the large model behind it are favored by more and more people.
2. GPT5 was suspended for "AI Dangerous Competition"
However, a special open letter signed by Musk and other 1,000 technologists took the lead in boycotting GPT5 to stop the "AI Dangerous Race".
They argue that AI labs and independent experts should jointly develop and implement a set of shared safety protocols for advanced AI design and development during the moratorium, subject to rigorous review and oversight by independent external experts.
When AI begins to align with or even exceed human capabilities, "possessing potential risks to society and humans" is no longer a sci-fi slogan.
GPT5 has been called to stop by the scientific and technological circles, is it useful? Once technology is developed, it is irreversible. If it is a semi-finished product, it is still possible to stop it. Now that it has been released, it is basically impossible to stop it, unless an administrative order is issued. Otherwise, this thing can only go forward, it is impossible to go backward.
Musk can do it himself. But he has no right to ask other companies not to do it, and they all know that this is a very dangerous thing.
3. Stopping is unstoppable
Once Pandora's Box is opened, it cannot be closed. . . Anyone who has seen the three-body problem knows that the road of technology is essentially a road of no return.
Moderation is fine, overdoing it is a dead end. Without moral control, there is no greater ability to control the world. Human civilization will be destroyed by humans. People are no longer people. It's just a meat worm under the control of technology. As human emotions, as human civilization and emotion. Will turn into a cloud of smoke in front of this cold procedure.
It brings people in the future, only money, only technology, and there will be no top art, such as calligraphy, Chinese painting, and poetry. this kind of thing.
It is convenient for people, but it seems to be busier. Human technology has evolved to this point, but happiness has become an extremely luxurious thing. Tragedy!
4. No rush to train GPT-5
OpenAI founder Sam Altman also came out to confirm that he is not in a hurry to train GPT-5, and he will not train in the short term.
Likely Reasons In the paper describing GPT-4, OpenAI predicts diminishing returns in scaling model size. There are also physical limitations on how many data centers OpenAI can build and how fast they can build them.
In addition to the reasons for the large model design and architecture itself, the huge cost is also a major reason. According to reports, ChatGPT needs at least 10,000 GPUs for training, and more resources are needed to ensure continuous operation. And Nvidia's latest H100 GPU designed for AI and high-performance computing has a unit price of $30,603, which is undoubtedly expensive. By Altman's own admission, GPT-4 cost more than $100 million to train.
OpenAI said that although GPT-5 is not being trained, it does not mean that the capabilities of GPT-4 have not been continuously expanded. “But we are doing other things on top of GPT-4, and I think there are various security issues that need to be addressed that were not addressed at all in the letter,” the CEO said.
Sam Altman's statement is interesting -- though not necessarily because it reveals OpenAI's future plans. Instead, they highlight a major challenge in the AI safety debate: the difficulty of measuring and tracking progress.
A large number of netizens translated Altman's words very badly, "This statement can be summed up as follows: we did not engage in GPT-5, but GPT-4.99999".
What do you guys think?
