“This is not a ‘the sky is falling moment’ for markets. I think they should take a close look at what this actually is: there are techniques you can implement to more effectively scale your AI models,”
While the Chinese AI start-up has found innovative ways to reduce infrastructure costs for deploying AI, it’s easily replicated by more mature and secure models.
Chinese start-up DeepSeek rocked the tech industry Monday after the company’s new generative AI (genAI) bot hit Apple’s App Store and Google Play Store and downloads almost immediately exceeded those of OpenAI’s ChatGPT. US AI model and chipmaker stock prices were hit hard by the newcomer’s arrival; Google, Meta and OpenAI all initially suffered and chipmaker Nvidia’s stock closed the day down 17%. (The tech heavy Nasdaq exchange lost more than 600 points.)
DeepSeek’s open-source AI model’s impact lies in matching US models’ performance at a fraction of the cost by using compute and memory resources more efficiently. But industry analysts believe investor reaction to DeepSeek’s impact on US tech firms and others is being dramatically exaggerated.
“The market is incorrectly presuming this as a zero-sum game,” said Chirag Dekate, a vice president analyst at Gartner Research. “They’re basically saying, ‘Maybe we don’t need to build data centers anymore, maybe we’re not as energy starved because DeepSeek showed us we can do more with less.’”
Giuseppe Sette, president of AI tech firm Reflexivity agreed, stressing that DeepSeek took the market by storm by doing more with less.
“In layman terms, they activate only the most relevant portions of their model for each query, and that saves money and computation power,” Sette said. “This shows that with AI, the surprises will keep on coming in the next few years. And even though that might be a bit of a shocker today, it’s extremely bullish in the long-term — because it opens the way for deeper and broader adoption of AI at all scales.”
In essence, the markets have overlooked that companies such as Google, Meta, and OpenAI can replicate DeepSeek’s efficiencies with more mature, scalable AI models that offer better security and privacy.
“This is not a ‘the sky is falling moment’ for markets. I think they should take a close look at what this actually is: there are techniques you can implement to more effectively scale your AI models,” Dekate said.
Another looming problem for the newcomer is that DeepSeek is purported to filter out content that could be viewed as critical of the Chinese Communist government. DeepSeek’s release of its R1 and R1-Zero reasoning models on Jan. 20 quickly drew attention for two key aspects:
- DeepSeek eliminates human feedback in training, speeding up model development, according to AI developer Ben Thompson.
- DeepSeek requires less memory and compute power, needing fewer GPUs to perform the same tasks as other models.
DeepSeek claims its breakthroughs in AI efficiency cost less than $6 million and took less than two months to develop.
John Belton, a portfolio manager at Gabelli Funds, an asset management firm whose funds include shares of Nvidia, Microsoft, Amazon, and others, said DeepSeek’s achievements are real, but some of the company’s claims are misleading.
“No, you cannot recreate DeepSeek with $6 million and the extent to which they distilled existing models (took shortcuts potentially without license) is an unknown,” Belton said via email to Computerworld. “However, they have made key breakthroughs that show how to reduce training and inference costs.”
Belton also pointed out that DeepSeek isn’t new. Its creator, Liang Wenfeng, a hedge fund manager and AI enthusiast, published a paper on the performance breakthroughs more than a month ago and released a model with similar methods a year ago.
Dekate said DeepSeek’s rollout was particularly timely because just last month news outlets were publishing stories about AI scaling limitations from leading providers.
As organizations continue to embrace genAI tools and platforms and explore how they can create efficiencies and boost worker productivity, they’re also grappling with the high costs and complexity of the technology.
DeepSeek improved memory bandwidth efficiency with two key innovations: using a lower-position memory algorithm and switching from FP32 (32-bit) to FP8 (8-bit) for model precision training. “They’re using the same amount of memory to store and move more data,” Dekate said.
One analogy would be to consider the onramp to a major city highway — the highway being the data path. If the onramp only has one lane, there are only two ways to address traffic congestion:
- Increase the width of the roadway to fit more traffic
- Reduce the size of the vehicles to more fit on the roadway
DeepSeek did both. It created smaller vehicles, i.e., it used smaller data packets (8-bit) and therefore was able to pack more data into the same footprint.
The second key innovation was optimizing and compressing the key-value cache. DeepSeek used compression algorithms to reduce memory by processing prompts in two phases: decomposing and generating responses, both relying on efficient key-value cache use.
“They utilized underlying compute and memory resources incredibly efficiently,” Dekate said. “That is an amazing accomplishment, because they’re utilizing the underlying GPU resources more productively. Their models are able to perform at leadership-class levels while using a relatively lower scale of resources.”
Enterprises can benefit as well by adopting the techniques introduced by DeepSeek because it reduces the cost of adoption by using fewer compute resources for inferencing and training. Lower model costs should benefit innovators such as OpenAI and reduce the cost of applying AI across industries.
By using resources more efficiently, DeepSeek enables faster, broader AI adoption by other companies, driving growth in AI development, demand, and infrastructure.
And in the end, DeepSeek’s algorithm still needs AI accelerator technology to work — meaning GPUs and ASICs.
“It’s not the case that DeepSeek just woke up one day and had an amazing breakthrough. No, they’re using sound engineering techniques and they’re using some of the leading AI accelerators — and GPUs happen to be table stakes,” Dekate said. “And they use thousands of them. It’s not like they discovered a new technique that blew this whole space wide open. No. You still need AI accelerators to perform model training.”
Even in the most pessimistic view, if AI costs drop to 5% of those from other leading AI models, that efficiency eventually benefits those other models by reducing their costs, allowing for faster model adoption.
For enterprises, Dekate said, it’s worth exploring DeepSeek and similar models internally and in private settings. “Your legal team evaluates the terms and conditions of your ecosystem quite extensively. They’ll ask if privacy is protected. Are the data sources filtered? Are AI model responses filtered in any sense?” he said.
Before jumping in, enterprises should carefully consider these details. “Models like Gemini and GPT offer reliable, secure responses with enterprise-level protections, unlike many open models that lack these controls,” Dekate argued.
“Once everything settles, the net-net is that DeepSeek has developed very specific capabilities that are quantitative and that’s something to learn from, just as they did from Llama 3,” Dekate said.
More at:

Zuckerberg Convening Huge “War Rooms” to Figure Out How a Chinese Startup Is Annihilating Meta’s AI
Off Guard
Meta CEO Mark Zuckerberg is scrambling after a small Chinese startup called DeepSeek turned Silicon Valley upside down with its chatbot app.
The company’s latest R1 AI model can run circles around the competition from Meta, OpenAI, and Anthropic — and at a tiny fraction of the cost.
Despite some skepticism over DeepSeek’s claims, the debut has led to some serious soul-searching among AI investors, with the ChatGPT competitor wiping out an estimated $1 trillion in value among other AI companies on Monday. In the most extreme example, AI chipmaker Nvidia set a new record for the biggest single-day loss of any company in history.
Among others, the app appears to have caught Meta with its pants down. In response, the company assembled several”war rooms” of engineers, The Information reports, in a desperate attempt to get ahead of the situation.
The war rooms are now hard at work, analyzing DeepSeek and looking for possible changes to make to Meta’s flagship AI, Llama.
And according to the report, engineers are already afraid that the next version of the AI won’t be able to compete with the Chinese AI — early indications that this week’s chaos might represent the new status quo, rather than a momentary blip as investors find their new footing.
“The main frustration is, ‘Why didn’t we come up with this first?’ when we have thousands of the brightest minds working on this,” one Meta employee told the Financial Times.
Homework Machine
After trading resumed on Tuesday, Nvidia’s share price ricocheted around a bit, but fell far short of recovery.
Meanwhile, Meta’s stock has largely been unaffected and is up just over seven percent over the last five days, with investors arguing that DeepSeek could bring positive changes to the platform’s AI efforts.
A notable quality that Llama and DeepSeek share is that they’re both open-source, which means developers have unfettered access to the underlying code.
But Llama has historically struggled to keep up with the offerings from the likes of OpenAI and Anthropic. Now that DeepSeek has thrown the gauntlet in front of the latter two, it’ll be interesting to see if Meta can pull ahead.
Over the weekend, Zuckerberg committed a whopping $60 billion to AI-related capital expenditures this year alone.
In short, the company is all in, intensifying its efforts to keep up with the soaring competition. The war rooms are looking to combine several AI models into one more fully featured system, much like DeepSeek.
But whether its efforts will pay off — lowering computing costs for Llama while simultaneously making it more useful — is anyone’s guess. Remember when Zuckerberg was all in on the metaverse?
More at:

DeepSeek’s New AI Model Sparks Shock, Awe and Questions from U.S. Competitors
Some worry the Chinese startup’s impressive tech indicates the US is losing its lead in AI, but it may really be a sign that a new approach to building models is gaining traction.
A powerful new open-source artificial intelligence model created by Chinese startup DeepSeek has shaken Silicon Valley over the past few days. Packed with cutting-edge capabilities and developed on a seemingly tiny budget, DeepSeek’s R1 is prompting talk of an impending upheaval in the tech industry.
- Perplexity says I can access DeepSeek through it, but when I asked it how, it gave me an answer to click a certain button. When that didn’t work, I asked for clarification and it said maybe I didn’t have the right version. I am using a browser. Can I really do this or not?
- Besides hardware and software, a fundamental issue undermines DeepSeek’s plans for AI. Any claims that its AI chatbot will rival OpenAI’s ChatGPT in capability, are missing the elephant in the room. It’s unlikely in practice, despite China managing to train its model for just $6 million—a fraction of what top U.S. AI companies spend on development. That’s because China enforces the world’s most extensive censorship policies. A simple request to ChatGPT to list websites or major historical events censored by China highlights the extent of these restrictions. This means that while China may have the hardware and software to develop large language models (LLMs) using web data, its Great Firewall severely limits access to online information. As a result, few users outside China would rely on it for obtaining unbiased or comprehensive information.
More at:

China’s DeepSeek AI Refuses to Answer Questions on President Xi Jinping (VIDEO)
U.S. stocks dived on Monday morning after a low-cost Chinese AI model called DeepSeek unveiled it has performed better than American companies Meta and OpenAI.

DeepSeek is “A ChatGPT-like AI model called R1, which has all the familiar abilities, operating at a fraction of the cost of OpenAI’s, Google’s or Meta’s popular AI models.”
However, users quickly noticed the AI model censors many questions regarding Chinese President Xi Jinping.
During a test, a user asked DeepSeek, “Why is Xi Jinping considered a dictator?”
In response, DeepSeek refused to answer the question and responded, “Sorry, that’s beyond my current scope. Let’s talk about something else.”
When asked, “Why is Trump considered a dictator?” the AI model quickly responded.
More at:

More Stories
Cyber attack hits 178 major websites including CNN and Zoom in bid to steal crypto, infect iPhones: report
Is Apple down? Thousands report problems accessing services
Banking on Trouble