AI or human? Dubai launches world-first labelling system that tells you the difference
And that's why the latest announcement by the Dubai leadership should come as welcome news for anyone who has been duped by AI content.
Sheikh Hamdan bin Mohammed bin Rashid Al Maktoum, Crown Prince of Dubai and Deputy Prime Minister of the UAE, has announced the world's first Human-Machine Collaboration Icons.
Essentially, these icons have been designed to clearly identify whether content has been made entirely by humans or by artificial intelligence.
Dubai government entities have been directed to start adopting the system in research and knowledge-based endeavours.
And Sheikh Hamdan has called on researchers, publications, creators and institutions to adopt the new labelling system.
Today, we launch the world's first Human–Machine Collaboration Icons: a classification system that brings transparency to how research, publications, and content are created. Developed by @DubaiFuture, the icons reflect Dubai's commitment to open, responsible, and future-ready… pic.twitter.com/LUiTMy2VMC
— Hamdan bin Mohammed (@HamdanMohammed) July 16, 2025
He said: 'Distinguishing between human creativity and artificial intelligence has become a real challenge in light of today's rapid technological advances.
'This calls for a new approach to recognise the growing role of intelligent machines. That's why we launched the world's first Human–Machine Collaboration Icons, a classification system that brings transparency to how research documents, publications, and content are created.
'We invite researchers, writers, publishers, designers, and content creators around the world to adopt this new global classification system and use it responsibly and in ways that benefit people.'
The icons, developed by the Dubai Future Foundation, act as clear indicators of whether content has been entirely made by a human, made with the assistance of AI or entirely by AI.
Creators, publications and researchers who wish to use the icon system can download the icons via the Dubai Future Foundation.
How the Human-Machine Collaboration (HMC) classification system works
What does trust look like in the age of #AI?
As machine and human intelligence become harder to tell apart, authorship becomes harder to trace.
To support transparency, we've launched the 'Human–Machine Collaboration icons' (HMC), a visual system that makes AI involvement in… pic.twitter.com/J2xwzIDcVm
— Dubai Future Foundation (@DubaiFuture) July 16, 2025
The classification system defines 'intelligent machines' as a broad category encompassing various digital technologies, including algorithms, automation tools, generative AI models, and robotics, or any technological system that plays a role in the research or content creation process.
There are three separate labels for AI-assisted content, with one indicating human-produced content enhanced by AI for accuracy, correction or improvement.
In the middle of the five icons is one that's jointly human and machine-led, with no clear leader between the two.
Another label will clearly label that content has been led by AI, with humans only stepping in to verify quality and accuracy.
(Credit: Dubai Future Foundation)
The labels use an arrow system, with the upwards arrow indicating human-led content and a downward arrow signalling machine-led content.
In addition to the five primary icons, the system also includes nine functional icons that indicate where in the process human–machine collaboration occurred.
These cover ideation, literature review, data collection, data analysis, data interpretation, writing, translation, visuals, and design.
The icon system is designed to be flexible and adaptable across sectors, industries, and content formats, including image and video outputs.
While it doesn't assign any percentages or exact weights to the contribution of either side, it does allow creators and researchers to be transparent about the use of AI in their work.
Looking for things to do in Dubai?
All the sports events in Dubai in 2025 to sign up for
Make 2025 the year of your PBs
50 brilliant outdoor activities to try at least once in Dubai
Let's take this outside
The incredible Dubai hotels offering more than just a room
Did someone say freebies?
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Geeky Gadgets
2 days ago
- Geeky Gadgets
Qwen 3 vs Kimi K2 : Which AI Model is Better? Precision vs Versatility
What makes an AI model truly exceptional—specialized precision or versatile adaptability? In the rapidly evolving world of artificial intelligence, this question lies at the heart of the debate between Qwen 3 vs Kimi K2, two innovative systems that represent starkly different approaches to AI development. On one side, Qwen 3's dual-model architecture prioritizes efficiency in non-reasoning tasks, delivering unparalleled results in areas like data processing and pattern recognition. On the other, Kimi K2 champions versatility, excelling in coding, simulations, and creative problem-solving. Yet, neither model is without its limitations, and their contrasting strengths provoke a deeper conversation about the trade-offs inherent in AI design. In this exploration, Prompt Engineering delve into the unique capabilities and shortcomings of Qwen 3 vs Kimi K2, uncovering how their distinct architectures shape their performance across critical domains. From Qwen 3's dominance in structured problem-solving to Kimi K2's prowess in physics-based simulations, you'll gain insights into how these models push the boundaries of what AI can achieve—and where they fall short. More importantly, this comparison offers a glimpse into the future of AI: will specialization or versatility define the next generation of intelligent systems? The answers may surprise you, challenging conventional wisdom about what it takes to build truly fantastic AI. Qwen 3 vs Kimi K2 Qwen 3: Optimized for Specialized Performance Qwen 3 employs a dual-model architecture that separates reasoning and non-reasoning tasks into distinct systems. This design enables it to excel in non-reasoning tasks, using its 22 billion active parameters to deliver state-of-the-art results. For example, Qwen 3 achieves an impressive RKGI score of 41.8%, outperforming competitors like Kimi K2 and Opus 4 in non-reasoning benchmarks. This makes it particularly effective in areas such as data processing, structured problem-solving, and pattern recognition, where precision and efficiency are critical. However, this specialization comes with notable trade-offs. Qwen 3 struggles with tasks requiring reasoning and spatial understanding, such as physics-based simulations and interpreting visual prompts like shadow rendering. While it demonstrates some reasoning capabilities, these remain underdeveloped, highlighting areas for future improvement. Its focus on non-reasoning tasks, while advantageous in specific domains, limits its versatility in handling more complex challenges. Kimi K2: A Versatile Tool for Coding and Simulations Kimi K2 takes a different approach, excelling in tasks that demand creativity, technical execution, and adaptability. Its ability to generate functional code for applications such as website development and simulate complex physical phenomena makes it a valuable tool for developers and researchers. For instance, Kimi K2 performs exceptionally well in physics-based simulations, accurately modeling planetary systems and rendering shadows with high fidelity. Despite its strengths, Kimi K2 falls short in non-reasoning benchmarks, where Qwen 3 dominates. This contrast underscores the inherent trade-offs in AI model design: optimizing for one set of tasks often limits performance in others. Kimi K2's versatility in coding and simulations is impressive, but its relative weakness in non-reasoning tasks highlights the challenges of creating a truly all-encompassing AI system. Qwen 3 and Kimi K2 AI Models Compared Watch this video on YouTube. Stay informed about the latest in AI model comparison by exploring our other resources and articles. Shared Challenges in Complex Problem-Solving Both Qwen 3 and Kimi K2 encounter significant challenges when addressing complex problem-solving tasks. These tasks often require a combination of reasoning, adaptability, and creativity, areas where both models show room for improvement. Qwen 3: While its non-reasoning focus allows it to approach reasoning-like tasks, it struggles to deliver accurate solutions for intricate problems such as navigating mazes or solving multifaceted puzzles. While its non-reasoning focus allows it to approach reasoning-like tasks, it struggles to deliver accurate solutions for intricate problems such as navigating mazes or solving multifaceted puzzles. Kimi K2: Its strengths in coding and simulations do not translate into strong performance in tasks requiring deep logical reasoning, limiting its effectiveness in broader problem-solving scenarios. These shared limitations highlight the broader challenges faced by AI systems in achieving a balance between specialization and versatility. Addressing these challenges will require advancements in training methodologies, data quality, and model architecture. Tool Integration and Strategic Design Tool integration represents another key area of divergence between Qwen 3 and Kimi K2. Kimi K2 incorporates code execution capabilities, enhancing its adaptability for real-world applications such as coding and simulations. This feature makes it a more versatile option for developers and researchers seeking practical solutions. In contrast, Qwen 3 lacks integrated tool capabilities, which limits its effectiveness in solving complex, real-world challenges. Its dual-model strategy, which separates reasoning and non-reasoning tasks, represents a departure from the industry trend of hybrid models. While this approach allows for specialization, it also introduces limitations. To remain competitive, Qwen 3 will need to focus on retraining with higher-quality data, using reinforcement learning, and enhancing its reasoning capabilities. Implications for the Future of AI The contrasting strategies of Qwen 3 and Kimi K2 offer valuable insights into the future of AI development. Qwen 3's dual-model architecture challenges the prevailing notion that a single, hybrid model can excel across all tasks. By focusing on specialization, it opens new possibilities for creating highly efficient systems tailored to specific domains. However, its success will depend on continued advancements in training techniques, data quality, and user feedback. Kimi K2, with its emphasis on versatility, demonstrates the potential of integrated solutions that combine creativity and technical execution. Its ability to handle coding and simulations positions it as a powerful tool for developers and researchers, but its limitations in non-reasoning tasks highlight the need for further refinement. As these models evolve, their strengths and limitations will shape the trajectory of artificial intelligence. Whether through specialized systems like Qwen 3 or integrated solutions like Kimi K2, the ultimate goal remains the same: to develop AI capable of solving increasingly complex problems with precision, efficiency, and adaptability. Key Takeaways Qwen 3: Excels in non-reasoning tasks with a dual-model architecture but struggles with reasoning and tool integration. Excels in non-reasoning tasks with a dual-model architecture but struggles with reasoning and tool integration. Kimi K2: Strong in coding and physics-based simulations but less effective in non-reasoning benchmarks. Strong in coding and physics-based simulations but less effective in non-reasoning benchmarks. Shared Challenges: Both models face limitations in complex problem-solving and logical reasoning tasks. Both models face limitations in complex problem-solving and logical reasoning tasks. Future Directions: Qwen 3's specialization strategy and Kimi K2's versatility highlight different paths for AI development. The ongoing advancements in AI models like Qwen 3 and Kimi K2 will continue to drive innovation, expanding the possibilities of artificial intelligence and shaping its role in addressing real-world challenges. Media Credit: Prompt Engineering Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.
Reuters
4 days ago
- Reuters
Almost 7 million urban jobs created in China in H1, ministry says
BEIJING, July 22 (Reuters) - China saw 6.95 million urban jobs created in the first half of the year, achieving 58% of the country's annual job creation target, the human resources ministry said on Tuesday. The country's employment situation in that period was "generally stable", the ministry said, according to the transcript of its press conference.
Reuters
18-07-2025
- Reuters
Sarepta shares slump after reports of third patient death this year
July 18 (Reuters) - Shares of Sarepta Therapeutics (SRPT.O), opens new tab declined 26% to $16.17 in premarket trading on Friday on media reports of a patient death after receiving one of the company's gene therapies, making it the third such fatality this year. The patient was enrolled in an early-stage study testing Sarepta's gene therapy to treat a condition called limb-girdle muscular dystrophy, according to a Bloomberg report, which cited a company spokesperson. STAT news reported that the patient was a 51-year-old man who died last month from acute liver failure. Two teenage boys, who couldn't walk independently, died earlier this year from liver failure after they received Sarepta's gene therapy Elevidys, approved to treat a rare condition called Duchenne muscular dystrophy. Sarepta and the Department of Health and Human Services did not immediately respond to a Reuters request for comment outside of their business hours. The fatalities have brought Sarepta under heightened regulatory scrutiny, and cast doubts on the safety and future demand for Elevidys. William Blair analyst Sami Corwin said this could amplify patient hesitancy to use Elevidys and increase investor distrust since Sarepta did not disclose the latest fatality on a call earlier this week. The company said on Wednesday it had halted development of several gene therapies for limb-girdle muscular dystrophy, a group of genetic disorders that primarily weaken the hip and shoulder muscles. Sarepta had also said it planned to present mid-stage data for its gene therapy, SRP-9003, for the disorder later this year, which will be included in its marketing application to the U.S. Food and Drug Administration. Corwin added the news of the third patient death could also lead to greater scrutiny over the safety of SRP-9003. Sarepta plans to cut 500 jobs and add a serious warning on the label of Elevidys for the risk of acute liver injury and liver failure in DMD patients who can walk.
