I don't know if you've heard by now, but today's dip is mainly because DeepSeek, a Chinese artificial intelligence company has developed an open-source artificial intelligence model that some say works better than ChatGPT and other US AIs, and more importantly, does it for a fraction of the cost. This comes shortly
after the US announced a $500B investment in AI.The tech bubble has been behind the success of Nvidia, the world's most profitable company until MicroStrategy came into play. These companies, many of which have AI projects, have caused a huge demand for hardware that with the success of DeepSeek is now in doubt as to whether it will continue to be needed.
As a result, Nvidia is currently down 17% in today's session, and the Nasdaq is down 3.32%. As you may have noticed it has affected the markets in general and also bitcoin and cryptocurrencies which are also trading down today.
Here is an article on the subject:
Why Is DeepSeek Sinking Nvidia Stock?What do you think about it?
I have a very cautious attitude towards Chinese solutions “unparalleled in the world”. As practice has shown, it is almost always either an exaggeration or a purposeful manipulation. No, I do not completely exclude the possibility that the DeepSeek model is not bad, but....
1. There are NO confirmed and proven indicators of its efficiency and low consumption of computational resources
2. The system works with data only up to October 2023, but it has not been trained beyond that.
3. The system does not work with online data
4. In working with it (I personally tested it), I did not notice any pronounced advantages in speed, or for example from practice - in writing better code ChatGPT or the same Phind.
5. The timing of the publication of the “super news” is strangely coincidental with Trump's statements about large investments in AI in the US.
In short, we are still watching how the system will develop and what will be known about it, as well as what is actually “under the hood”.