DeepSeek vs ChatGPT in UAE: Will new AI model spark price wars?
UAE residents reported that DeepSeek delivers responses which feel more natural and exhibit fewer repetitions in contrast to ChatGPT. A large-scale cyberattack prevented some UAE residents from downloading DeepSeek on Monday.
This week DeepSeek became an international news topic after announcing that its AI models function with less data while demonstrating capabilities to compete against American AI companies.
DeepSeek built its chatbot with just $5.6 million and it reached the top position on Apple’s US app store which sent shockwaves through the market and triggered stock sell-offs for tech giants Nvidia, Microsoft, and Meta. OpenAI CEO Sam Altman called it an “impressive model” while former US President Donald Trump described it as "a wakeup call for our industries".
My experience with ChatGPT spans several months during which its performance has proven dependable for daily activities. The testing of DeepSeek revealed distinct characteristics that caught my attention. For starters, DeepSeek’s responses feel more natural.
ChatGPT often adopts an 'AI tone' which is known for its helpful yet formulaic responses. DeepSeek eliminates the mechanical feel present in other models. The answer sequence feels more natural and conversational according to student Iffah Sheeraz from Ajman.
“The 'Deepthink R1' feature is a standout. The feature allows users to examine the AI’s thought process and understand the steps it takes to reach its conclusions. Transparent explanations increase usefulness when addressing complicated inquiries. The medical student revealed that while ChatGPT provides answers it doesn't show the underlying methodology.
Sheeraz observed that ChatGPT sometimes uses repetitive phrasing but DeepSeek creates more diverse responses which maintains user interest.
Impact of cyberattack
As a Dubai resident interested in technological advancements that streamline work processes Kathrynne Nicole Alday discovered DeepSeek during her AI tool research and found it to be knowledgeable and accurate while efficiently handling complex tasks with clear explanations.
The tool caught my attention so I wanted to check whether it would enhance my work processes and strategic approaches. The registration on the platform was unavailable due to large-scale malicious attacks which caused delays in service as I tried to sign up to explore DeepSeek's offerings. Please wait and try again'," she said.
Alday reported a three-hour delay in the delivery of the authentication code.
The 22-year-old resident acknowledged that these issues are natural for growing platforms but remains hopeful about the platform's potential to improve professional efficiency and creativity.
Democratising AI
Jacob Falkencrone from Saxo Bank stated that DeepSeek's ascent has disrupted the AI investment sector because its large language model (LLM) performs on par with top US models such as OpenAI but costs only five to ten percent of their price.
Investors express concern regarding the expenses involved in AI infrastructure development amid ongoing price conflicts. Although current market responses create some worry investors should maintain confidence because AI's future prospects remain strong.
DeepSeek’s affordable AI models can democratise artificial intelligence technology and enable smaller innovators to participate in the market. The development of AI continues according to him indicating that its journey has not ended but is progressing.
Falkencrone identifies cost-lowering and accessibility-expanding innovations as drivers of the upcoming growth wave.
Investors need to maintain their composure while expanding their portfolio and keeping their long-term goals in mind.
Too early to assess
Julius Baer digital assets analyst Manuel Villegas says it is premature to evaluate whether DeepSeek represents “a true game changer” in AI infrastructure development due to its performance, cost-effectiveness, technological framework and scalability potential.
DeepSeek achieved innovation by training a complex model with Mixture of Experts (MoE) architecture at what seems to be an extremely low cost. According to Villegas the western approach has prioritized increasing chip counts over architectural optimization.
Unanswered questions
The senior analyst from Swissquote Bank Ipek Ozkardeskaya pointed out several unresolved queries about DeepSeek's potential integration with other applications and whether its construction cost remained below $6 million while American AI models demonstrated price tags in the hundreds of millions.
DeepSeek appears to have replicated an existing product at lower production costs. Their final product did not bring forth any new invention. The success of DeepSeek in delivering high-performance AI solutions at lower costs would benefit Chinese domestic firms and global businesses with tight budgets as they adopt AI technologies into their daily operations.
Comments