By : Bakinam Khaled
DeepSeek, a rising AI startup from China, has quickly captured the attention of social media influencers with its R1 model, an advanced large language model (LLM) that challenges industry leaders like OpenAI and Anthropic at a fraction of the cost. Its efficiency in reasoning tasks, open-source framework, and minimal reliance on high-end GPUs have fueled discussions, triggering market reactions and raising questions about AI’s evolving competitive landscape, reveals the Social Media Analytics Platform of GlobalData, a leading data and analytics company.
Influencers are highlighting DeepSeek R1's impressive performance in reducing reliance on high-end GPUs. The sudden impact on tech stocks, particularly Nvidia, a major GPU supplier, is also a significant focus of influencer discussion, with some viewing it as an overreaction.
Shreyasee Majumder, Social Media Analyst at GlobalData, comments: “The discussions among the influencers highlight DeepSeek R1's significant cost savings compared to existing LLMs. R1's inference cost is approximately 93% lower than its predecessor (o1), enabling local execution on high-end workstations and eliminating rate limits. This efficiency stems from the algorithmic breakthroughs, including training in FP8 and multi-token prediction, allowing a significantly low million training cost. However, this figure omits prior research expenses, implying that replicating DeepSeek's success requires substantial pre-existing investment.
“The rapid ascent of DeepSeek to the top of the US App Store has further intensified speculation regarding its success, raising questions about whether it stems from genuine technological innovation or a broader geopolitical strategy. Additionally, significant concerns have emerged regarding data privacy and potential national security risks, particularly related to the possible transfer of US user data to China.”
Below are a few popular influencer opinions captured by GlobalData’s Social Media Analytics Platform:
David O. Sacks, Founder & Partner at Craft Ventures:
“DeepSeek R1 shows that the AI race will be very competitive and that President Trump was right to rescind the Biden EO, which hamstrung American AI companies without asking whether China would do the same. (Obviously not.) I’m confident in the U.S. but we can’t be complacent.”
Gary Marcus, Professor of Psychology and Neural Science at New York University:
“NVidia is having a really bad day. OpenAI’s day is worse. Their business model (charging big bucks for LLMs) basically just blew up, DeepSeek is running circles around them, and a lot of the remaining talent is probably going to start thinking about whether to jump ship. And all this after Elon called BS on Stargate, while Microsoft largely opted out. NVidia will definitely still be here next year; OpenAI could conceivably run out of cash.”
Marc Benioff, Chair & CEO at Salesforce:
“Deepseek is now #1 on the AppStore, surpassing ChatGPT—no NVIDIA supercomputers or $100M needed. The real treasure of AI isn’t the UI or the model—they’ve become commodities. The true value lies in data and metadata, the oxygen fueling AI’s potential. The future’s fortune? It’s in our data. Deepgold.”
Gavin Baker, CIO & Managing Partner at Atreides Management:
“1) DeepSeek r1 is real with important nuances. Most important is the fact that r1 is so much cheaper and more efficient to inference than o1, not from the $6m training figure. r1 costs 93% less to *use* than o1 per each API, can be run locally on a high end work station and does not seem to have hit any rate limits which is wild. Simple math is that every 1b active parameters requires 1 gb of RAM in FP8, so r1 requires 37 gb of RAM. Batching massively lowers costs and more compute increases tokens/second so still advantages to inference in the cloud. Would also note that there are true geopolitical dynamics at play here and I don’t think it is a coincidence that this came out right after “Stargate.” RIP, $500 billion - we hardly even knew you….”
Daniel Ives, Global Head of Technology Research at Wedbush Securities:
“DeepSeek is a competitive LLM model for consumer use cases…launching broader AI infrastructure a whole other ballgame and nothing with DeepSeek makes us believe anything different. It’s about AGI for Big Tech and DeepSeek’s noise. Also no US tech using this tech. Buying oppy”
Avichal Garg, Managing Partner at Electric Capital:
“Since this has gotten some traction: - The DeepSeek team is obviously very talented - The DS tech is amazing and impressive - No hate intended at them or the tech I remain skeptical of the narrative, just like I don’t believe stories about how famous companies were founded and don’t believe the US Govt’s stated narratives on a whole host of issues. Foundational models are intertwined with geopolitics. When governments and geopolitics are involved, being skeptical is a good starting position.”
Ross Gerber, CEO at Gerber Kawasaki Wealth & Investment Management:
“If you trust what the Chinese tell you. If you put this app on your phone it has access to your information. Read the privacy policy. No thank you. #DeepSeek”
