DeepSeek vs ChatGpt Which one is better for Students?

DeepSeek vs ChatGpt

 Which one is better for Students?




In my previous blog, “The Dragon Is Out” I showed how Deepseek's emergence has significantly dropped the cost of Computing in AI models and how medium to small size companies can now build their AI models from Scratch by downloading the complete AI model of Deepseek on their local computers. Even a small Rasberry Pie ( A device created for Students) can handle Deepseek offline.

Now that's for developers and programmers what is in it for general users like me and you in DeepSeek and how is it different from ChatGPT?

 

Deepseek R.1 model is different in a way that it thinks and reasons more like we humans do so what it does whenever you ask a question there is this small icon right at the bottom of Deepseek (R.1) click this when you pose a question it will start the process of thinking breaking it down in smaller parts and then arguing and then answering it like humans

Riddle Test DeepSeek

When I gave a simple riddle to its R.1 model that's an answer it came up with.

 



In Deepseek, we can see that it gives a detailed answer after arguing with itself. It also analyzes the riddle by stating, "The question cleverly plays on the common assumption that it refers to a month only having 28 days. 😊” Ending it with a smiley Emoji


Riddle ChatGpt.

Posing the same riddle to ChatGPT and see what it came up with interestingly after DeepSeek, open AI added a small button like deepSeek at the bottom of its AI Model as well terming it reason, copying DeepSeek but is it as good as DeepSeek lets click on it and see.


As you can see ChatGpt also provides the correct answer but its reasoning is only a one-liner “Thought about months with 28 days for a second” Then showing the thought process and just came up with the answer not analyzing the riddle properly


The Winner

The Winner in a Simple Riddle is DeepSeek it provides a comprehensive answer and explains why the question confused its audience.

 



Creativity Test Stroy Telling DeepSeek.

I asked DeepSeek to tell me the story of RumpleStilSkin and it came up with this.




The Story is brief yet beautifully written using a classic style similar to the original tale.

 

Creativity Test Stroy Telling ChatGpt.

Now I gave the same story to ChatGpt Let's see what it does with it.




The story is retold in a nice style sticking to its original tale.

Which One is better?

I am a fan of JK Rowling's Harry Potter so I will prefer the deepSeek version of the Story as much better retold simply because of the creative liberty taken in it plus the lyrical tune and the background story beautifully added.

While testing it out in ChatGpt I put Both the Stories in ChatGpt Labelling ChatGpt version No.1 and labeled Deepseek story No.2 that's what it came up with

As you can see it agrees with me and rejects its own generated version of the story.


Conclusion

Deepseek although new in town surpasses Open AI ChatGpt model on many counts. It's R.1 model which reasons before answering is a game changer.

Hats off to the team of DeepSeek who put this extraordinary model on top of Open AI. The Dominance of the USA in AI is dented severely for sure.

Since I am into reading and writing more, I use both models to analyze and research for students like me having options like DeepSeek which reasons and then answers is a far better choice than ChatGpt.

Comments

Popular posts from this blog

Bubbly Bunch's Day Out

From Hogwarts to History | How to Chat with Any Book Character

How to overcome Procastination.