In a stunning turn of events, DeepSeek, a trailblazing AI company from China, captured the tech industry’s attention with their remarkable breakthrough. Using their R1 V3 model, which cost significantly less to develop, they outperformed OpenAI’s exclusive o1 reasoning model in multiple areas, such as programming, mathematics, and scientific problem-solving.
The appearance of the model led to a significant and abrupt decline in stock prices within the market. Notably, NVIDIA experienced a drop of approximately $600 billion in market capitalization, which could potentially be the most substantial single-day decrease ever recorded. Despite DeepSeek’s enigmatic success, Meta’s Chief Technology Officer, Andrew Bosworth, asserts that he anticipated these events six months prior to their occurrence, based on reports from Business Insider.
In response to an Instagram follower’s query regarding the origin of DeepSeek, Bosworth playfully stated that it was quite amusing to observe its development.
Six months ago, I had already sent an email predicting that something similar to DeepSeek would emerge, without knowing exactly what. Thus, it wasn’t as shocking for those of us in this field, compared to those who are not.
According to reports, Meta allegedly tried to make up ground in the AI competition by employing rough tactics such as teaching their AI systems using copyrighted materials, while also implementing strategies to hide any potential copyright violations.
We had been following DeepSeek as it was recently launched, and soon after, it started making headlines. I believe it’s significant yet not as sensationalized as portrayed. The launch of DeepSeek is notable due to its innovative open-source nature. They have indeed made remarkable strides in memory architectures for model construction, pushing the boundaries in reasoning models. It seems they have also made substantial improvements by refining their work against existing models, significantly advancing the field.
DeepSeek is a “great thing, but it’s not a world-changing thing,” concluded Bosworth.
It’s true that DeepSeek’s arrival has sparked some surprise within the industry. An email from Microsoft CFO Amy Hood advised employees to keep their focus on company objectives amidst the latest AI breakthroughs, possibly including OpenAI’s $500 billion commitment to construct advanced AI systems called Stargate. However, Microsoft CEO Satya Nadella expressed admiration for DeepSeek, stating that it could be beneficial for the business.
In other locations, it’s been proposed that DeepSeek invested approximately $1.6 billion and utilized as many as 50,000 AI-focused NVIDIA graphics processing units in the development of its open-source model. This runs counter to assertions about its economic efficiency.
Read More
- DUSK PREDICTION. DUSK cryptocurrency
- LDO PREDICTION. LDO cryptocurrency
- DGB PREDICTION. DGB cryptocurrency
- Blockchain is the best fintech to ensure Sharia ethics — Web3 exec
- EUR INR PREDICTION
- Marvel Rivals Season 1 is bringing the Fantastic Four — here’s the Invisible Woman in action
- Looks Like DOOM 64 Is Getting a Native PS5 Port
- JTO PREDICTION. JTO cryptocurrency
- TRB PREDICTION. TRB cryptocurrency
- Welcome Home
2025-02-07 15:09