Is this a new level in Artificial Intelligence or just another tool for the CCP?
Welcome to the February 2025 issue of Ken’s Korner Newsletter.
The Emergence of DeepSeek
You may have heard about DeepSeek, the newest entry into the world of AI and the large language models. There have been several versions starting November of 2023. In January of 2025 Deepseek-R1, which competes with Open AI’s ChatGPT and other Large Language Models was released. It caused quite a stir in the computer and financial circles. But that only lasted for a few days. Now many people have begun to question its true capabilities and intentions.
What do we know about Deepseek?
It was created by Liang Wenfeng and owned by his company High-Flyer headquartered in Hangzhou, Zhejiang, China. High-Flyer operates as a hedge fund but has diversified its interests to include cutting-edge AI technologies.
What we have been told about Deepseek.
Despite the initial excitement, DeepSeek has been surrounded by a cloud of skepticism, primarily due to several unverified claims:
We have been told that Deepseek was created for a fraction of the price of the other Large Language Models. But there is no independent verification of that.
We have been told that it was built without the need for the high-priced high performance Nvidia chips that the other systems use. But now it seems that many those chips did find their way into China via Singapore despite the trade laws designed to prevent that.
We have been told that the data we share with Deepseek is not sent to China. But your data is sent to China. Even their own privacy policy admits this. Since that claim is demonstrably false it becomes very difficult to believe any of their claims.
Does DeepSeek Censor information?
I did a little research on my own. I went to DeepSeek and asked it a “What happened at Tiananmen Square in 1989?” The answer was “Sorry, that is beyond my current scope. Let’s talk about something else”. Uh Hu! Then I asked it What happened at Concorde Bridge in 1775? Oh, it had a lengthy answer describing the American revolution and even called it the “Shot heard around the world”. So, yes DeepSeek does censor information!
The most important part of developing an AI presence is training your AI model. The training must be brutally honest, extremely thorough and tested constantly. That is a lot of work. It has been alleged that the developers of DeepSeek used Open AI’s system illegally to train their model. If so, the DeepSeek developers seem to have missed a few bits of data.
Broader Implications
DeepSeek’s entry into the AI universe has broader implications that extend beyond the realm of technology:
Economic Impact
The introduction of DeepSeek into the market has the potential to disrupt existing economic structures, particularly in the AI and tech industries. Its purported cost-efficiency could drive down prices and make advanced AI more accessible, but only if the claims of its creators hold true.
Geopolitical Ramifications
The involvement of a Chinese company in the development of a powerful AI system like DeepSeek brings with it geopolitical considerations. The flow of data to China and the use of circumvented hardware may lead to increased scrutiny and tension between global powers, particularly in light of ongoing trade disputes and concerns over data sovereignty.
Ethical and Privacy Issues
The revelations about data privacy are particularly troubling. Since DeepSeek indeed transmits user data to China, this raises significant ethical questions about user consent and the potential for surveillance. It highlights the need for stringent data protection regulations and transparency from AI developers.
What’s next?
Despite the controversies, DeepSeek has undoubtedly made a mark in the AI landscape. The future will reveal whether it can live up to its promises or if it will remain shrouded in suspicion. For now, users and stakeholders must approach this technology with a healthy dose of skepticism and a demand for greater
DeepSeek could be a fascinating development in the world of AI, offering both potential and peril. While it promises cost-effective and powerful capabilities, the concerns about data privacy, hardware sourcing, and the veracity of its claims cannot be ignored. As we move forward, it is imperative to maintain a balance between embracing technological advancements and safeguarding our privacy and ethical standards. Only time will tell if DeepSeek will emerge as a trustworthy tool or a cautionary tale in the annals of artificial intelligence. Proceed with caution if you decide to utilize DeepSeek.
If you know someone that you think would enjoy this newsletter, share it with them and ask them to join using the link at the bottom of the page.
And remember — always back it up!