Be patient. Let the R&D process unfold. It's a technical issue amongst peers.
Not politicians and journalists balling and pouting and name calling on TV.
Google Mountain View CA came out w LLM Open Source in 2014. R&D people started using LLM then.
Many teams have made improvements to the Open Source.
Not just the LLM, but also the legacy Source Codes: Python3, Anaconada 3, Data Science, Sci Kit, Machine Learning, Tensor Flow, Py Torch, OpenCV, etc.
Deep Seek made the improvements proprietary.
That's their perogative.
Other teams are reverse engineering the improvements now and will update the Open Source.
Then the R&D process repeats on and on.
Deep Seek made their modifications to the source code proprietary, but they released marketing papers and engineering details of what they offer. Thus, World Wide ,teams of engineers are parsing the words for what Deep Seek has done. Improvements will be encoded into the Google LLM Source Codes. It won't result in an exact copy. But enough to get a Ph.D Candidate though his thesis.
Gotta love that ai nonsense is running the market. Imagine the entire tech industry built on hype of a product that does absolutely nothing. Here we go.
First, market commentators and investors seem to be believing, without the slightest bit of skepticism, the words of a Chinese hedge fund manager that this DeepSeek product only took two months and $6 million to create. It is FAR more likely that it took years and hundreds of billions from the Chinese government.
Second, the same people are spazzing over some paranoid MAGA programmer geek's "Sputnik moment" quote. That guy was just waiting for his moment of X/Twit fame to unleash that exaggeration. He played it well, but doesn't mean it's anywhere near true.
Third, this is not tested, and in fact they butchered the rollout with fake news about being hacked, when the real problem they couldn't fix their registration system. Does not bode well.
Fourth, it doesn't matter how many downloads they were able to do. There is no way a Chinese AI system is going to have significant market impact on the AI market in North America and Europe over the next five years. A year from now, most of us will be "DeepSeek who?"
The wise amongst us will stock up on all the AI related stocks that tanked today (including NVDA). AI is not going anywhere and DeepSeek ain't sh!t.
First, market commentators and investors seem to be believing, without the slightest bit of skepticism, the words of a Chinese hedge fund manager that this DeepSeek product only took two months and $6 million to create. It is FAR more likely that it took years and hundreds of billions from the Chinese government.
Second, the same people are spazzing over some paranoid MAGA programmer geek's "Sputnik moment" quote. That guy was just waiting for his moment of X/Twit fame to unleash that exaggeration. He played it well, but doesn't mean it's anywhere near true.
Third, this is not tested, and in fact they butchered the rollout with fake news about being hacked, when the real problem they couldn't fix their registration system. Does not bode well.
Fourth, it doesn't matter how many downloads they were able to do. There is no way a Chinese AI system is going to have significant market impact on the AI market in North America and Europe over the next five years. A year from now, most of us will be "DeepSeek who?"
The wise amongst us will stock up on all the AI related stocks that tanked today (including NVDA). AI is not going anywhere and DeepSeek ain't sh!t.
Right. So you're saying "buy the dip!"? Very clever, AND original!
DeepSeek-R1 matches the capabilities of leading AI models like OpenAI’s ChatGPT but was developed at a fraction of the cost. This efficiency challenges the prevailing notion that advanced AI development requires substantial financial and computational resources.
Unlike many proprietary models, DeepSeek-R1’s code and technical details are openly available. This transparency allows developers worldwide to access, modify, and integrate the model into various applications, fostering innovation and collaboration. It’s going to be used more than the others for sure.
Also, following its release, DeepSeek-R1 quickly surpassed ChatGPT as the top free app in the U.S. Apple App Store. Its user-friendly interface and free access have contributed to its widespread adoption.
The success of DeepSeek-R1 has led to significant shifts in the tech industry, including a notable decline in the stock prices of major companies like Nvidia and Microsoft. This development has prompted discussions about the future dynamics of the AI market and the future impact on tech stocks.
There are also geopolitical implications. DeepSeek’s advancements have raised questions about the global AI race, particularly concerning U.S. export controls and China’s potential to set international AI standards. The company’s achievements suggest a shifting landscape in AI development and deployment.
DeepSeek’s combination of high performance, cost efficiency, open-source transparency, and rapid market penetration positions it as a transformative force in the AI industry. It will be interesting how US companies respond.
It is a bit reminiscent of compact disc players. I remember when they first hit the market. They were only at audiophile stores and cost $700-800 (probably $2,500ish adjusted for inflation). About ten years later, you could by a portable CD player for about $40 at Kmart.
DeepSeek is pretty much the $40 portable CD player to NVIDIA's $800 early market entrant. It wasn't supposed to move this fast. The big AI companies were supposed to be light years ahead of any small time market entrant because of the restrictions on selling advanced chips to China and the short term memory of Silicon Valley in forgetting how often the tech giants get their buts beat by tiny start ups.
You are truly the Cliff Clavin of LR. DeepSeek-R1 is an LLM. It is apparently a competitor to OpenAI, Llama, etc. Nvidia is a hardware company. It makes the chips that all of the AI companies use. DeepSeek used older Nvidia chips for training its model.
DeepSeek-R1 matches the capabilities of leading AI models like OpenAI’s ChatGPT but was developed at a fraction of the cost. This efficiency challenges the prevailing notion that advanced AI development requires substantial financial and computational resources.
Unlike many proprietary models, DeepSeek-R1’s code and technical details are openly available. This transparency allows developers worldwide to access, modify, and integrate the model into various applications, fostering innovation and collaboration. It’s going to be used more than the others for sure.
Also, following its release, DeepSeek-R1 quickly surpassed ChatGPT as the top free app in the U.S. Apple App Store. Its user-friendly interface and free access have contributed to its widespread adoption.
The success of DeepSeek-R1 has led to significant shifts in the tech industry, including a notable decline in the stock prices of major companies like Nvidia and Microsoft. This development has prompted discussions about the future dynamics of the AI market and the future impact on tech stocks.
There are also geopolitical implications. DeepSeek’s advancements have raised questions about the global AI race, particularly concerning U.S. export controls and China’s potential to set international AI standards. The company’s achievements suggest a shifting landscape in AI development and deployment.
DeepSeek’s combination of high performance, cost efficiency, open-source transparency, and rapid market penetration positions it as a transformative force in the AI industry. It will be interesting how US companies respond.
Nice to know exactly what ChatGPT knows about DeepSeek.
Thet claim they built the system for $5M. Any good Linux R&D A.I. Software Engineer can build Google LLM from github, tweak certain parts and call it a new A.I. system. It's really isn't all new. Its 99% inherited Open Source. Even the Linux Kernel, Linux Utilities have been rewritten (pilfered?) from AT&T SVr4, Berkeley UNIX, Sun Microsystems Solaris, HPUX, IBM AIX, CMU Mach UNIX X-Windows, etc. from the 70s/80s. You don't start from scratch. The Chinese on the project I presume inherited Chinese Mandarin, they probably did I don't know them. I know they did not invent the fresh water that flushed the toilet in their office.
First, market commentators and investors seem to be believing, without the slightest bit of skepticism, the words of a Chinese hedge fund manager that this DeepSeek product only took two months and $6 million to create. It is FAR more likely that it took years and hundreds of billions from the Chinese government.
Second, the same people are spazzing over some paranoid MAGA programmer geek's "Sputnik moment" quote. That guy was just waiting for his moment of X/Twit fame to unleash that exaggeration. He played it well, but doesn't mean it's anywhere near true.
Third, this is not tested, and in fact they butchered the rollout with fake news about being hacked, when the real problem they couldn't fix their registration system. Does not bode well.
Fourth, it doesn't matter how many downloads they were able to do. There is no way a Chinese AI system is going to have significant market impact on the AI market in North America and Europe over the next five years. A year from now, most of us will be "DeepSeek who?"
The wise amongst us will stock up on all the AI related stocks that tanked today (including NVDA). AI is not going anywhere and DeepSeek ain't sh!t.
Ah, yes, the “wise amongst us,” stockpiling AI stocks like canned beans in a doomsday bunker because clearly DeepSeek is just a figment of China’s imagination and won’t amount to anything. Sure, it’s far more plausible that Beijing emptied its piggy bank on a multi-billion-dollar secret project, rather than a scrappy startup achieving efficient results. And hey, what’s a massive wave of downloads and global attention when it doesn’t come with your personal seal of approval?
Also, bravo for dismissing the “Sputnik moment” while simultaneously hoping the very companies hit by DeepSeek’s rise will magically bounce back without any disruption. The cognitive dissonance is dazzling. But don’t worry—when everyone else is asking “DeepSeek who?” next year, I’m sure you’ll be busy enjoying your NVIDIA stock gains… unless, of course, you’re too busy eating crow.
Second, the same people are spazzing over some paranoid MAGA programmer geek's "Sputnik moment" quote.
Deep Seek is not quite a "Sputnik Moment" I would say. My great-grandpa worked at North American Aviation in Santa Monica, famous for the X1 Charles Yeager 1st Supersonic jet, an X15 Neil Armstrong 1st Outer Space Jet. My great-grandpa was a Vegetable Buyer and Produce Man in South Central L.A. when Sputnik was launched. He had a Mech E from the G.I. Bill but there was no jobs for Nisei Engineers in those days. But by the end of the week, every Mech E in California had the job they wanted in Defense. It made the childhood for my grandparents quite nice.
