The founder of the Black Myth Wukong commented on DeepSeek: It may be a scientific and technological achievement at the national level
According to online reports, DeepSeek has caused heated discussions recently. The founder of the black myth Wukong commented on DeepSeek: It may be a scientific and technological achievement at the national level. The following is the full text: Let's seriously push DeepSeek before the year. I have been using V3 for a month, but R1 has only been used for 5 days, so I have to post this Weibo post. Let me start with a violent theory: DeepSeek may be a scientific and technological achievement at the national level. In order to explain how amazing this achievement is, let me make an analogy: If there is a large AI model that achieves any of the following, it will be a super breakthrough- 1. Powerful. Compared with O1's reasoning ability, it's not as good as O1 for the time being. 2. It's cheap. With fewer parameters, training overhead and usage cost are an order of magnitude smaller. 3. Open source. Anyone can download and deploy it themselves, provide papers detailing training steps and tips, and even provide mini models that can run on mobile phones. 4. Free. The services currently provided by the official government are completely free and can be used by anyone anytime and anywhere. 5. Internet access. For the time being, the only reasoning model that supports online search (o1 does not support it yet) 6. Native. Deep Quest is a very small-scale young China company. A local team with no overseas experience or even senior experience developed and completed the above six points. DeepSeek did all and at the same time. It's too 6, 6 is not true. If this is not worth cheering for, what is there to cheer for? So, what should you do now when you see this Weibo post? 1. Please visit directly and use it immediately. There are also mobile apps. 2. Use it hard and crazily, and try to basically replace traditional search with it. Imagine you already know the top experts in every field, and they are all good friends you are online at all times. Ask it first-whether it's building a complex function, figuring out which nearby swimming pool is best, checking legal traps in the contract, getting newly written lyrics to rhyme a little more, or helping parents confirm whether a health care product really works. 3. Go see how others use it, try other big models, Understand what AI is good at, what it is not good at, and how to train it, and then continue to unlock and iterate on your own usages and more tools. I hope DeepSeek R1 will disfascinate you with the most advanced AI and gradually turn AI into water and electricity in your life.
Disclaimer: The views in this article are from the original Creator and do not represent the views or position of Hawk Insight. The content of the article is for reference, communication and learning only, and does not constitute investment advice. If it involves copyright issues, please contact us for deletion.