In an attempt to solve this mystery, when experts asked questions to this AI, some interesting things came to light. When this model was asked who made it, he refused to tell clearly. But he definitely told that he has been trained in Chinese language and he has information about the world till May 2025. This date is very important because China’s famous AI company DeepSeek also has almost the same information. For this reason, people are speculating that perhaps this is a new and secret project of DeepSeek. However, neither OpenRouter nor DeepSeek has given any official statement in this regard.
What is Hunter Alpha’s power?
You can estimate the power of Hunter Alpha from the fact that it has 1 trillion parameters. In simple language, parameters mean the ability of the brain with which it understands and responds to language. 1 trillion means 10 trillion, which is a very huge number. Apart from this, its biggest quality is its memory. It can read and understand up to 1 million words (1 million tokens) at a time. That means if you give it a thick book, it will read it in a jiffy and answer every question you have about it.
It is estimated that Google’s model Gemini Ultra also works on 1 trillion parameters. Similarly, if we talk about GPT 4, there are 1.8 trillion parameters. Meta’s Llama 3 only works with 405 billion parameters, while Grok 3 only works with about 300 billion parameters.
AI engineer Nabil Haouam says, “If a model with such a huge memory and tremendous power of thinking is available for free, then it is really a big and special thing. Usually, one has to pay a lot of money for a model with such features.”
Why doubt DeepSeek only?
Now a debate has erupted among experts around the world whether this is really DeepSeek’s next big model V4? There were already reports in China that DeepSeek is preparing to launch its new and updated version. The working method of Hunter Alpha is also very similar to DeepSeek.
AI expert Daniel Dewhurst believes, “Any AI has a special way of thinking, which is difficult to hide. Hunter Alpha’s way of reasoning is exactly the same as we have seen in earlier DeepSeek models.” However, some people also disagree with this. Independent AI tester Umur Ozkul says that there are some minor differences in its design, which make it different from DeepSeek.
Why is the company hiding its name?
Hiding names has become a new strategy by companies. Companies launch new models in the market without disclosing their names, so that people can use them without any bias and the company can know how well their model is actually working. Earlier, a similar model named ‘Pony Alpha’ had also come, which was later claimed by a big company as its own.
A warning has also been given regarding the use of Hunter Alpha. It is clearly written in its profile that whatever you talk to it or whatever information you give to it, it can be recorded so that this model can be further improved. Despite this, there is huge enthusiasm about it among developers and computer engineers. Within a few days of its launch, it has processed more than 160 billion words. People are using it to create new software and complete complex tasks. At present, the whole world is waiting for the mystery to be revealed as to who is the real owner of this Hunter Alpha.
Discover more from News Link360
Subscribe to get the latest posts sent to your email.
