Posted on 01/29/2025 6:01:18 AM PST by delta7
It was only a matter of time before an innovative mind created the next mainstream AI tool to compete with ChatGPT. In a massive step toward AI advancement, Liang Wenfeng of China launched DeepSeek, an open-source large language models (LLM) intended to compete if not one day overshadow ChatGPT. The launch immediately wiped $1 trillion off the US stock exchange and the tech competition between China and the US is coming to a head.
ChatGPT is run by OpenAI. Its creation marked the dawn of a new way of interacting with the internet and accessing information. Users can ask AI to instantaneously perform actions and it is reshaping the way the world operated. People have created businesses based on ChatGPT.
There have been countless warnings of AI replacing human jobs. Governments are still uncertain how to regulate these services and the data they pull from users. Of course, countless services like ChatGPT have launched in recent years, but DeepSeek may be the next best alternative.
Wenfeng hired all the top minds graduating from Chinese universities and paid them top dollar to create DeepSeek for a fraction of what it took to create ChatGPT. OpenAI’s GPT-4, launched in 2023, cost $100 million to develop; DeepSeek-R1 began with a $6 million investment.
ChatGPT write own code
Semiconductor chips are shaping the tech race. It takes semiconductor chips to operate these AI programs and that has been an ongoing problem for American companies. Chinese companies do not have such problems.
There is much speculation that ChatGPT did not require the estimated 10,000 GPUs and 3,500 NVIDIA servers. Nonetheless, DeepSeek is operating on less. Now, DeepSeek has around 50,000 NVIDIA H100 chips but they cannot speak about the matter due to US export controls.
President Trump stated that DeepSeek is a reminder that American companies need to be “laser focused” on competing with China. Everyone is impressed at the low operating costs. “Instead of spending billions and billions, you’ll spend less, and you’ll come up with, hopefully, the same solution,” Trump noted. Some US politicians are already calling for a ban. “The U.S. cannot allow CCP models such as DeepSeek to risk our national security and leverage our technology to advance their AI ambitions.
We must work to swiftly place stronger export controls on technologies critical to DeepSeek’s AI infrastructure,” Rep. John Moolenaar, R-Mich., the chair of the House Select Committee on China, said Monday.
Hangzhou DeepSeek Artificial Intelligence Co., Ltd, owned and funded by hedge fund High-Flyer has inserted $2 trillion into the US markets at the time of this writing. Critics claim that DeepSeek censors available information based on what the CCP will and will not permit. Still, investors seem extremely bullish on DeepSeek, which has already surpassed ChatGPT as the most downloaded AI app on the Apple app store.
Moneyback marked as a troll retread and fully Zot! worthy.
Posts own blog to try to smear others.
This is a totally stupid statement. Politicians evidently don't know what Open Source means. It means that the source code used to develop the application is available to anyone, usually free of charge. I was a computer science prof for 10 years at a Big Ten university so I know what Open Source means and the benefits it affords the development community. For US software developers to follow the advice offered here and not stand of the shoulders of those who have gone before us is an ostrich approach to AI development. Someone should take Rep. Moolenaar aside and explain to his what Open Source is all about.
Why is it assumed that any AI program is superior to natural intelligence?
They knocked 6% off nasdaq by following this chinese stock on one day.
Alexa, tell me how to win!
Listen up, Kamala,....
The technology isn’t comprehensive.
Smells like Ozone.
Here kitty kitty.
Quick, but incomplete responses often have considerable value.
No human knows most everything about most everything.
Copilot on ‘global warming’:
Global warming is a significant and pressing issue that refers to the long-term increase in Earth’s average surface temperature due to human activities, primarily the emission of greenhouse gases like carbon dioxide (CO2) and methane (CH4). These gases trap heat in the atmosphere, leading to a range of environmental impacts, including:
Rising sea levels due to melting polar ice caps and glaciers.
More frequent and severe weather events, such as hurricanes, droughts, and heatwaves.
Disruption of ecosystems and loss of biodiversity.
Ocean acidification, which affects marine life.
Addressing global warming requires a collective effort to reduce greenhouse gas emissions, transition to renewable energy sources, and implement sustainable practices. It’s a complex challenge, but every small action can contribute to a larger positive impact.
I think the answer is speed. AI is not a substitute for human intelligence. Rather, it is a means for quickly organizing data to provide an answer to a question. AI cannot guarantee that its answer is right, or the best solution to a given problem. The dollar loss by the NASDAC is an overreaction to the DeepSeek announcement, especially in light of its Open Source position.
The US has about 100 million bovines and about 3 million square miles.
Do you really expect me to believe one belching bull or cow per 20 acres could impact the climate?
“Some US politicians are already calling for a ban.”
An acquaintance at Stanford has told me DeepSeek is banned from the electronics of all electronic devices of employees and students.
Does anyone else see how shallow this guys understanding of how computers and AI works? Is the underlined part of the quote above a surprise to anyone over the age of 5?
Well, Kamala thinks it all happens in clouds.
The $5.6 million development cost seems wildly low compared to the vast resources typically required to train a competitive LLM. Even if China has cheaper labor and state-backed infrastructure, AI training is still hardware-intensive and requires massive compute power—which isn’t cheap.
Other red flags:
Speed of Development (55 days claimed) – AI models take years to refine, especially at ChatGPT-level quality. Even with shortcuts, getting top-tier performance fast is highly unlikely.
Energy Efficiency Claims – AI training is extremely energy-intensive. If DeepSeek is significantly more efficient than GPT-4, it would raise serious questions about what corners were cut.
H800 GPUs – U.S. export controls restrict China’s access to NVIDIA H100 chips, forcing them to rely on H800s, a downgraded version. If DeepSeek was trained on these, performance may be inferior to Western models.
China’s AI Warfare & Propaganda – China plays the long game in tech wars, often using state subsidies, data control, and strategic deception to appear ahead. They have incentives to inflate their AI capabilities to shift global perception and policy.
Bottom line: If DeepSeek is as powerful, efficient, and cost-effective as claimed, either China has made an unprecedented AI breakthrough or they’re bluffing. Given the stakes, deception is a real possibility.
Does anyone else see how shallow this guys understanding of how computers and AI works? Is the underlined part of the quote above a surprise to anyone over the age of 5?I do. Many other people do, too. I don't need to even read this article. See my reply #2. As usual, this guy really doesn't understand what he is writing about.
“Why is it assumed that any AI program is superior to natural intelligence?”
I have a large yard. I have bushes that need pruning and weeds to be pulled. No machine right now would let me type away inside my house while tiding up my yard.
If you want to get really concerned, read your gmail license. They not only can read your private emails, they reserve the right to use that information use that information as they see fit. They also have reserved the right to sell your information to whomever they wish and you have no recourse. I avoid Google products as much as I can.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.