We analysed the online hype around DeepSeek and found something unusual: thousands of fake accounts were amplifying the narrative of mass adoption.
This wasn’t just random spam — the accounts showed patterns typical of coordinated bot networks (synchronous posting, recycled avatars, and disproportionate engagement).
The result: investors and the market briefly reacted to a level of “traction” that wasn’t real.
I’m curious how others here think about this:
How can we distinguish genuine user adoption from manufactured buzz in an LLM/AI market that is moving this fast?
What tools or heuristics do you use to check the authenticity of online signals?
(Research details and case breakdown in the post.)
This reads like a propaganda piece, there's lots of bullet points and hardly any real information to support your bold claim. From the few examples you do show, they all look pretty standard. Not something that would have an impact... unless DeepSeek really was a great success and really did change the AI landscape... which it did.
Surely investors and tech leaders, as distinct from speculators, should be testing the product and doing due diligence on the company. I'm just a layman and I liked how a small version of deepseek ran faster and with less power draw on my PC than other small version models did. Output, very informally tested, was good. I did not think the 'reasoning' text added anything useful but I'm also not qualified to say how it contributed to the final output.
> I'm just a layman and I liked how a small version of deepseek ran faster and with less power draw on my PC than other small version models did.
This doesn't sound accurate. There are no "small versions" of deepseek r1. They released some distilled fine-tunes of their big model. They took qwen2.5 models as a starting point, and fine-tuned them with ~800k generations made with deepseek r1. But they did not change any of the model's architecture, so the "base" model released by qwen and the fine-tuned one released by deepseek should run identically on your hardware.
Yes, that's ollama's fault. They renamed the fine-tuned qwen model to deepseek, and a lot of people got confused. But behind the naming blunder, the model you got was qwen2.5-finetuned-r1, which was identical in architecture with qwen2.5 (so it should have ran at the exact same speed)
Folks, this is what bad stats look like. All bot accounts look and behave like this. They didn't even do a systematic study to prove their claims against other bot acconts (which are not associated with deepseek at all).
I don't see any network analysis on this page. What network analysis do you see?
I do see generic statements like "boosting each other", and I see vaguely-drawn lines in the primary diagram with no further explanation, but that hardly counts as network analysis, right?
"Fake accounts" like OP's (evai) who somehow registered in January 2021 and its only activity is this single post today and some AI generated comment to go along with it? Definitely something suspicious but I'm not sure it's DeepSeek here.
“ 3,388 were fake accounts — about 15% of all engagement, double the usual baseline”
Double the usual baseline of what? Bots spamming the most popular #OfTheDay?
nobody is immune to propaganda, but this one slid off me like water off of ducks back
I have no horse in the AI race or in the US-China rivalry, but at the time the hype felt weird. There was so much gloating that rubbed me the wrong way, and frankly, I notice that in plenty of China-related discussions. I always wonder: Why would Westerners, whom I suspect frequent the sites I visit the most, frolic so much that China is #1? Last time I was befuddled was during the India-Pakistan scramble. The comments were only focusing on how a Chinese jet downed a French jet and I thought to myself: There are two nuclear powers going at it and all they talk about is China? The French, too, later claimed that this was a concerted effort by bots.
We analysed the online hype around DeepSeek and found something unusual: thousands of fake accounts were amplifying the narrative of mass adoption.
This wasn’t just random spam — the accounts showed patterns typical of coordinated bot networks (synchronous posting, recycled avatars, and disproportionate engagement).
The result: investors and the market briefly reacted to a level of “traction” that wasn’t real.
I’m curious how others here think about this:
How can we distinguish genuine user adoption from manufactured buzz in an LLM/AI market that is moving this fast?
What tools or heuristics do you use to check the authenticity of online signals?
(Research details and case breakdown in the post.)
This reads like a propaganda piece, there's lots of bullet points and hardly any real information to support your bold claim. From the few examples you do show, they all look pretty standard. Not something that would have an impact... unless DeepSeek really was a great success and really did change the AI landscape... which it did.
Surely investors and tech leaders, as distinct from speculators, should be testing the product and doing due diligence on the company. I'm just a layman and I liked how a small version of deepseek ran faster and with less power draw on my PC than other small version models did. Output, very informally tested, was good. I did not think the 'reasoning' text added anything useful but I'm also not qualified to say how it contributed to the final output.
> I'm just a layman and I liked how a small version of deepseek ran faster and with less power draw on my PC than other small version models did.
This doesn't sound accurate. There are no "small versions" of deepseek r1. They released some distilled fine-tunes of their big model. They took qwen2.5 models as a starting point, and fine-tuned them with ~800k generations made with deepseek r1. But they did not change any of the model's architecture, so the "base" model released by qwen and the fine-tuned one released by deepseek should run identically on your hardware.
Emphasis on layman and I'm not at my PC but when deepseek came out there was something I could pull in ollama that could run on modest hardware.
Yes, that's ollama's fault. They renamed the fine-tuned qwen model to deepseek, and a lot of people got confused. But behind the naming blunder, the model you got was qwen2.5-finetuned-r1, which was identical in architecture with qwen2.5 (so it should have ran at the exact same speed)
Folks, this is what bad stats look like. All bot accounts look and behave like this. They didn't even do a systematic study to prove their claims against other bot acconts (which are not associated with deepseek at all).
This is network analysis which is used to identify networks of interest.
I don't see any network analysis on this page. What network analysis do you see?
I do see generic statements like "boosting each other", and I see vaguely-drawn lines in the primary diagram with no further explanation, but that hardly counts as network analysis, right?
"Fake accounts" like OP's (evai) who somehow registered in January 2021 and its only activity is this single post today and some AI generated comment to go along with it? Definitely something suspicious but I'm not sure it's DeepSeek here.
Markets have been vulnerable to bollocks since time began. Bots are even older
“ 3,388 were fake accounts — about 15% of all engagement, double the usual baseline” Double the usual baseline of what? Bots spamming the most popular #OfTheDay?
nobody is immune to propaganda, but this one slid off me like water off of ducks back
I doubt that the WSJ is influenced by a couple of thousands of bot accounts that are overtly spammy (there is also #Solana spam in the comments):
https://www.wsj.com/tech/ai/china-ai-deepseek-chatbot-6ac4ad...
https://www.wsj.com/tech/ai/china-deepseek-ai-nvidia-openai-...
https://www.wsj.com/tech/ai/deepseek-ai-how-it-works-725cb46...
The Western "AI" influencers and shills are much more sophisticated than that.
What a nonsense article, I encourage people to look at the example tweets they cite.
I have no horse in the AI race or in the US-China rivalry, but at the time the hype felt weird. There was so much gloating that rubbed me the wrong way, and frankly, I notice that in plenty of China-related discussions. I always wonder: Why would Westerners, whom I suspect frequent the sites I visit the most, frolic so much that China is #1? Last time I was befuddled was during the India-Pakistan scramble. The comments were only focusing on how a Chinese jet downed a French jet and I thought to myself: There are two nuclear powers going at it and all they talk about is China? The French, too, later claimed that this was a concerted effort by bots.
Seems like another disinformation misinformation
[dead]