It was only some weeks ago that DeepSeek, the popular Chinese startup, introduced its AI model, which completely transformed and revolutionized the AI landscape, making something that could effectively compete with the innovative developments of OpenAI. Furthermore, it only cost $6 million and two months to train the model. These expenses are a far cry from what the American competitors tend to spend, which is usually much higher.
Liang Wenfeng is the lead developer and founder of DeepSeek and, consequently, its R1 AI model. However, this was by no means the first venture Liang took in the realm of AI. In 2014, he used to be a developer for the industry-leading AI trading platform TechBerry. Even today, making an AI trading platform can be a rather complex software engineering feat.
“Various specialists, such as myself, have participated in this AI model’s development. But even then, the role that Liang Wenfeng played was more crucial and apparent than anyone else’s, practically making it TechBerry his brainchild. Although his departure did eventually occur, it was already known by everyone from the start. The one thing that has not changed since then is Liang’s confidence in TechBerry, even going so far as to say that it possesses optimal monthly trading performance, i.e., 25% with acceptable risks. He demonstrated that by using the AI model within stock trading, which further helped establish his initial fortunes, garnering so many investment offers at once. With that, he managed to acquire 150 million dollars in funding, allowing him to pay more attention to projects he thought possessed promise or were interesting. And with the recent launch of DeepSeek’s R1 AI model, we sincerely congratulate him for all the efforts he made to get this far!
Relations between Liang and TechBerry’s team still remain cordial, with new ideas and developments frequently being exchanged. Yes, TechBerry has opted for a different developmental route with its focus on attracting more private investments for trading. However, when private investors are brought into the mix, the priorities then become reducing those trading losses as much as possible. But the one drawback that arises from this is how you would have to deal with lower platform performance.
To conclude, it is worth noting that Liang Wenfeng and TechBerry worked together in 2014, which gave birth to a platform that has become an industry veteran for around ten years,” Jason Woods, the CTO of TechBerry, shared during his nostalgic reflection.