E-Paper
Sign In
DeepSeek vs ChatGPT in UAE: Will new AI model spark price wars?
DeepSeek made headlines globally after it claimed that its models use less data and have the ability to challenge and match American AI companies
- PUBLISHED: Wed 29 Jan 2025, 6:00 AM UPDATED: Wed 29 Jan 2025, 8:33 PM
UAE residents have found China's AI sensation DeepSeek to provide responses that "feel more natural" and less repetitive, unlike its rival ChatGPT. However, some residents could not download it on Monday after it was hit with a large-scale cyberattack.
DeepSeek made headlines globally this week after it claimed that its models use less data and reportedly have the ability to challenge and match American artificial intelligence (AI) companies.
Recommended For You
DeepSeek said it spent only $5.6 million to build its chatbot which became the top-rated app on Apple's US app store, rattling the market and causing sell-off of giants like Nvidia, Microsoft, and Meta. Sam Altman, CEO of OpenAI, termed it an “impressive model” while US President Donald Trump said it's "a wakeup call for our industries".
Follow KT on WhatsApp Channels.
“I've been using ChatGPT for months and it's been reliable for everyday tasks. But after trying DeepSeek, I have noticed some interesting differences. For starters, DeepSeek's responses feel more natural. ChatGPT sometimes leans into that 'AI tone' we are all familiar with - helpful but formulaic. DeepSeek, on the other hand, avoids that robotic edge. It's subtle, but the flow of answers feels closer to a human conversation,” said Ajman-based student, Iffah Sheeraz.
“The 'Deepthink R1' feature is a standout. It lets you peek into the AI's reasoning process, showing how it arrives at conclusions. That transparency adds value, especially for complex questions. With ChatGPT, you get the final answer, but not the 'how' behind it,” said the 18-year-old medical student.
Sheeraz noted that ChatGPT occasionally repeats phrases or structures while DeepSeek's responses vary more, which keeps things fresh.
Sophia Matveeva, CEO and founder of Tech for Non-Techies, said its reasoning capacity is on par with ChatGPT and Claude, but it is currently experiencing server issues, which makes it hard to use.
“I decided to use ChatGPT, Claude and DeepSeek to do the same task to compare how they work," she said. The task was to analyse course reports from 60 students in her tech for non-technical founders course.
“First, DeepSeek had login issues, so I could not log in. This is normal when there is an influx of users, but still annoying. When I did manage to log in, I was pleasantly surprised. The interface is super easy to use and resembles the US LLMs. I wrote three prompts to help it understand who I am and what task I need it to do. It came back with good questions to my prompts. When I asked it to give me a structure for my report, the result was thoughtful and useful. It was just as good as Claude and ChatGPT,” said Matveeva.
However, she said it stopped working after her third prompt and said: "The server is busy. Please try again later.”
While it was frustrating, the issue was "solvable", she said. “For now, I have gone back to ChatGPT/ Claude to finish working on the report.”
However, once DeepSeek's server issues are solved, she said she would cancel her ChatGPT subscription and make the switch.
Impact of cyberattack
Dubai resident Kathrynne Nicole Alday, who has a strong interest in how innovations can make work easier and faster, said she recently came across DeepSeek while researching AI tools and said it was more knowledgeable, accurate, and capable of handling complex tasks with proper explanations.
“This caught my attention and I was curious to see if it could help improve my workflow and strategies. When I tried to sign up to explore the platform, I came across a message saying, 'Due to large-scale malicious attacks on DeepSeek's services, registration may be busy. Please wait and try again'," she said.
Alday explained that there was also a delay in receiving the authentication code, which took about three hours.
“While it was a bit inconvenient, I understand that such issues can happen with platforms gaining interest,” said the 22-year-old resident, adding that she is looking forward to seeing if it can truly deliver on its promises and help professionals like her work more efficiently and creatively.
DeepSeek is open source
Andreas Hassellöf, CEO of Ombori, said the advancements from DeepSeek were drastically reducing the cost of deploying advanced AI capabilities as these models would accelerate the adoption of AI in real-world applications, from smart cities and retail to healthcare and beyond.
“This efficiency is particularly crucial in the current landscape, where recent updates to US export controls on certain AI hardware have created tiers of access for different countries. The UAE, classified as a Tier 2 country, faces limitations in accessing critical hardware resources,” said Hassellöf.
However, with technologies like DeepSeek, fewer GPUs (graphics processing unit) are required to achieve the same, or even better, results - making the need for advanced hardware less pressing, he said.
“This can open up to a much broader range of countries, encouraging collaboration across borders and creating a more globally interconnected AI ecosystem. The fact that DeepSeek is open source will further amplify its impact, enabling developers worldwide to build on this foundation and drive innovation at an unprecedented pace,” said Hassellöf.
Democratising AI
Jacob Falkencrone, chief investment strategist at Saxo Bank, said DeepSeek's rise has sent shockwaves through the AI investment landscape as its large language model (LLM) matches the performance of US leaders like OpenAI - at just five to ten per cent of the cost.
“Investors are worried about AI infrastructure spending and price wars. But while the short-term market reaction seems concerning, the long-term outlook for AI remains robust. Cheaper models like DeepSeek's could democratise AI, opening doors for smaller players to innovate. In short, the AI story isn't over - it's just evolving,” he said.
Falkencrone sees innovations that lower costs and broaden accessibility sparking the next wave of growth.
“For investors, the key is to stay calm, diversify, and focus on the bigger picture.”
Too early to assess
Manuel Villegas, digital assets analyst at Julius Baer, believes it is too early to assess if DeepSeek is “a true game changer” and threat for its performance, cost-effectiveness, technologically-advanced framework, and scalability as they all could have a major impact on the development of AI infrastructure.
“The real innovation is that DeepSeek managed to implement and train, at an allegedly very low cost, a highly complex model that leverages the Mixture of Experts (MoE) architecture. The west has focused on essentially throwing more chips at it, instead of optimising the architecture,” said Villegas.
Unanswered questions
Ipek Ozkardeskaya, senior analyst at Swissquote Bank, said there are a lot of unanswered questions, including whether the DeepSeek model could be integrated and used by other apps and whether the company really built a model for less than $6 million whereas the price mark of the US AI models reached several hundred million dollars.
“Last but not least, DeepSeek looks like it made something that already existed for a cheaper price. However, it did not come up with an end product that did not exist. But if DeepSeek successfully does what it says it does - bring equal performance AI models for a cheaper price - it will help Chinese local players, and all-sized companies around the world that have limited budgets to integrate AI models into their daily lives,” she said.
ALSO READ:
DeepSeek hit by large-scale cyberattack as users flock to AI startup; registrations limited
Cryptos, stocks plunge amid selloff sparked by Deep Seek
Trump warns of 'wake-up call' as low-cost Chinese AI jolts sector
