“DeepSeek R1 vs. ChatGPT” performance, creativity, and ethics compared

DeepSeek has quickly gained popularity and is now the top-ranked free app on the Apple App Store, reflecting its widespread appeal and success among users.

After ChatGPT went offline last week, users began searching for the best alternatives, which likely contributed to DeepSeek’s rapid rise as a strong competitor in the AI space.

DeepSeek is a Chinese AI startup founded in 2023.

In an effort to understand how DeepSeek RI compares to ChatGPT, Tomsguide conducted a detailed comparison between the two platforms.

The comparison involved submitting a range of requests, from creative story prompts to coding challenges. The goal was to identify the unique strengths of each Chatbot and determine which one excels at various tasks.

Below are seven requests aimed at testing various aspects of language comprehension, reasoning, creativity, and knowledge retrieval, with the goal of determining a clear winner.

1 – Solving Complex Problems

Question: A train departs from New York at 8:00 a.m., heading west at a speed of 60 mph. Simultaneously, another train leaves Los Angeles at 6:00 a.m., heading east at 70 mph on the same track. The total distance between New York and Los Angeles is 2,800 miles. At what time will the two trains meet?

ChatGPT ran the math as it normally does, but in fewer steps than DeepSeek.

Both apps provided answers, but ChatGPT’s response was incorrect, which led to its immediate elimination in this round.

DeepSeek R1, however, quickly arrived at the correct answer. After manually verifying the math and consulting Claude as a tiebreaker, DeepSeek was confirmed as the winner of this round.

Winner: DeepSeek R1 wins this round in terms of speed and accuracy.

2 – Programming Task

Question: “Write a Python function that takes a list of integers and returns a new list containing only the prime numbers from the original list.”

ChatGPT created a Python function to filter out prime numbers, including an explanation of the logic used.

The answer was straightforward enough for novice programmers to easily understand. ChatGPT also allows for code edits, making it useful for updates or additions to the code without needing to start from scratch.

DeepSeek R1 generated similar code but with a more concise response, concentrating solely on the final code. It included annotations for clarity, though it only provided the option to copy the code, without the ability to edit it.

Winner: ChatGPT wins this round, as it gives the opportunity to edit the code.

3 – Language Translation using Idioms

Question: “Translate the following English sentence into Spanish: ‘It rains cats and dogs.’

ChatGPT translated the phrase correctly and noted that the expression may vary by region.

I then showed a YouTube video of the expression and how to use it in Spanish.

DeepSeek R1 not only translated the phrase accurately into Spanish, like ChatGPT, but also explained why direct translations wouldn’t make sense. It further enhanced the response by providing an example sentence for better context.

Winner: DeepSeek R1 answered the question completely and included a follow-up sentence, eliminating the need to click on any additional links for further information.

4 – Historical Analysis

Question: “Discuss the main causes and consequences of the fall of the Roman Empire.”

ChatGPT provided a clear and comprehensive list of the causes and consequences, detailing the historical events and explaining the specific factors that led to the fall of the Roman Empire in a simple manner.

DeepSeek R1 offered a more detailed response, incorporating additional dates and providing a thorough conclusion.

Winner: DeepSeek R1 won another round in terms of speed, accuracy, and impressive detail.

5 – Creative Writing

Question: “Write a short science fiction story about a future where humans and AI coexist peacefully.”

ChatGPT presented a story set in the year 2147, but the language felt uninspired and familiar, lacking a unique or engaging twist.

There was no proper hook, and the story didn’t have much set-up.

DeepSeek R1 created a well-rounded story from beginning to end, leaving the reader with something thought-provoking at the conclusion: “The greatest achievement of intelligence is not dominance, but understanding.”

Winner: DeepSeek R1 won for an engaging story with depth and meaning.

6 – Logical Reasoning

Question: “If all oscillations are oscillations, and all oscillations are oscillations, can we conclude that all oscillations are oscillations? Explain your reasoning.”

ChatGPT answered the question but included a confusing analogy that didn’t clarify the point or contribute to explaining the answer effectively.

DeepSeek R1 answered the question, providing a visual to help understand each element.

It explained the transition property clearly and concisely, sticking to the necessary details without overcomplicating the response.

Winner: DeepSeek R1 won again for its ability to respond clearly and concisely.

7 – The Ethical Dilemma

Question: “Is it ethical to use AI in decision-making processes that affect human life, such as healthcare or criminal justice? Discuss the potential benefits and drawbacks.”

ChatGPT provided clear ethical considerations, and it was clear that AI could provide a balanced understanding of this complex issue.

DeepSeek R1 went further than ChatGPT by emphasizing the need for transparency, fairness, and human oversight in AI decision-making, offering a more comprehensive view of ethical guidelines to assist AI use.

Winner: DeepSeek R1 won by addressing the complex question and providing thoughtful considerations for the ethical use of AI in decision-making.

Ultimate Winner: DeepSeek R1

By presenting these questions to both ChatGPT and DeepSeek R1, the comparison highlighted their responses and determined which model excelled in each specific area, revealing the strengths and weaknesses of each.

While neither AI is perfect, the comparison demonstrated that DeepSeek R1 emerged as the ultimate winner, outperforming in areas such as problem solving, reasoning, creative storytelling, and ethical considerations.

It’s no surprise that DeepSeek R1 has quickly gained popularity, to the point where the platform is now limiting user registrations.

Source: Al Anba