All this accusation of chatgpt of the Deepseek copy of Openai and Microsoft reminds me of a lesson that I have learned in the past 15 years writing on this kind of thing – most people do not care if something is a Copier. They will go for the thing that is cheaper and better.
No one cares that the Xiaomi SU7 EV is a clear scam of the Porsche Taycan. It is much cheaper, a little faster and has a longer range. For a consumer, the choice is obvious (hypothetically speaking, taking into account shipping restrictions to the United States).
Bringing him closer to his home, nobody hit an eyelid at Aldi to copy large brands products and sell them at lower costs under their own brands, but it is one of the most growth grocery stores Rapid from the country and the products are often selected in blind taste tests.
This does not mean that it is a perfect formula (looking at you, Instagram wires and each laptop that tries to overthrow the M3 MacBook Air), but the principle is solid. When something is cheaper, it is a real factor in the purchase decision of a person who crushes the originality most of the time.
I raise these examples to make a very simple point. People have two choices: or a system that you may have to pay (Chatgpt), or a free AI assistant who turned out to be the equivalent of a model that costs up to $ 200 per month.
I don’t know about you, but for most people in the future, the decision is obvious.
Do more with
So let’s make a land line here. Deepseek R1 all blind on Monday – the Chinese manufacturing AI model remains at the top of the Apple App Store list, because the company was able to form this assistant for $ 6 million (a fraction of billions of American people to companies of AI) using more chips and produce faster and more precise results.
Not only that, but the reasoning part of this model is much clearer to show you its process of reflection on the equivalent of Openai in O1. This later caused a kracée of billions of dollars on the stock market, Nvidia being the worst blow (although he rebounds now) and President Trump to say that it is an “alarm clock” for the IA industry of America.
In our Face-Off between Chatgpt and DeepseekThe writer Ai Amanda Casswell saw R1 reign supreme in “everything, problem solving and creative narration reasoning and ethical solutions”. Our comparison was with Chatgpt more by the way ($ 20 per month).
After that, Deepseek presented Janus – A model of image generation which is (to say the least) a little rough on the edges of our own tests, but it will progress over time.
And of course, there is a jank to that. Since Deepseek is a company belonging to Chinese, there is an annoying censorship problem that you can get around if you are creative, and there are major security problems concerning the Chinese government.
But the botter of all this is that Deepseek is open source. Users can download it and use it by themselves, look at their bowels and see how it works and fold it to their own requirements.
Meanwhile, Chatgpt is closed – which means you cannot get the underlying code. And while at the beginning, CEO Sam Altman praised Deepseek on his page X, OPENAI and Microsoft now study if the data is out of the depth formed by Chatgpt.
Openai’s activities are at risk
So, in the last three days, we have a Copycat AI assistant who is better and completely free to use. Companies care about it, the American government cares, but consumers will not do so because of their point of view, which will be “I do not want to spend $ 20 to help my homework, I will use the one which is free. “
In addition, Deepseek could be a massive risk for other Openai activities in the granting of its API API of AI to other companies. Reminder (not that Reminder) is a nice little tool that I use to collect the content that I see day by day on the Internet, summarize, find connections between everything and even ask questions to all this information to use the graph of knowledge for quick responses multiple facets.
According to OPENAI API documentationThe cost of using its technology is broken down into the amount of tokens used and can extend from $ 100 to $ 200,000 per month depending on the amount of your application used.
Would a company pay a lot to use chatgpt in its own applications continue to do so or spend path Less and use Deepseek R1? It is open source, after all, and open source technology will always be cheaper to use for businesses than to have to dismiss a closed source technology.
And let’s entertain the scenario that the US government prohibits Deepseek (like what was tempted with Tiktok). It is there and people can take this model and use it by themselves-host it on their own servers where the Chinese government cannot be heard.
By the way, none of these factors even in the other major risks of President Trump, introducing prices on tokens from Taiwan. Most industry experts predict It will considerably increase computer prices.
But more specific to Openai and Meta which are in this AI race, these prices could make the GPUs super powerful from Nvidia which manage their models much more expensive. This will actively prevent US companies from being competitive.
Don’t be bitter, be better
It brings me back to something my grandfather said when I was a child: “Don’t be bitter, be better.” Yes, it seems to have been torn directly a Pinterest inspiring table, but I think the point is true here.
AI’s arms race has seen a new competitor entering the fray of nowhere – Deepseek R1 has reached parity with Openai in the most important way.
Whether it is this feat by legitimately forming the model from zero or distilling the knowledge of the OpenAi models is not relevant, because (and I repeat) most people do not care No copies, they care about the price.
It’s time to be better.