Development

DeepSeek’s efficiency forces OpenAI to rethink closed AI model strategy

OpenAI has released reasoning-focused open-weight models in a strategic response to China’s surging AI ecosystem, led by DeepSeek’s disruptive efficiency. Unlike earlier coverage, the shift is framed not merely as competitive posturing but as a deeper recognition of shifting innovation philosophies.

DeepSeek’s rise stems from maximizing limited resources under the US’s export restrictions, proving that top-tier AI doesn’t require massive chip clusters. The agility has emboldened the open-source AI sector in China, where over 10 labs now rival those in the US, fundamentally reshaping competitive dynamics.

OpenAI’s ‘gpt-oss’ models, which reveal numerical parameters for customization, mark a departure from its traditional closed approach. Industry watchers see this as a hybrid play, retaining proprietary strengths while embracing openness to appeal to global developers.

The implications stretch beyond technology into geopolitics. US export controls may have inadvertently fueled Chinese AI innovation, with DeepSeek’s self-reliant architecture now serving as a proof point for resilience. DeepSeek’s achievement challenges the US’s historically resource-intensive approach to AI.

AI rivalry may spur collaboration or escalate competition. DeepSeek advances models like DeepSeek-MoE, while OpenAI strikes a balance between openness and monetization. Global AI dynamics shift, raising both technological and philosophical stakes.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Altman warns of harmful AI use after model backlash

OpenAI chief executive Sam Altman has warned that many ChatGPT users are engaging with AI in self-destructive ways. His comments follow backlash over the sudden discontinuation of GPT-4o and other older models, which he admitted was a mistake.

Altman said that users form powerful attachments to specific AI models, and while most can distinguish between reality and fiction, a small minority cannot. He stressed OpenAI’s responsibility to manage the risks for those in mentally fragile states.

Using ChatGPT as a therapist or life coach was not his concern, as many people already benefit from it. Instead, he worried about cases where advice subtly undermines a user’s long-term well-being.

The model removals triggered a huge social-media outcry, with complaints that newer versions offered shorter, less emotionally rich responses. OpenAI has since restored GPT-4o for Plus subscribers, while free users will only have access to GPT-5.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

New Instagram Map lets users share location with consent

Instagram has introduced an opt-in feature called Instagram Map, allowing users in the US to share their recent active location and explore location-based content.

Adam Mosseri, head of Instagram, clarified that location sharing is off by default and visible only when users choose to share.

Confusion arose as some users mistakenly believed their location was automatically shared because they could see themselves on the map upon opening the app.

The feature also displays location tags from Stories or Reels, making location-based content easier to find.

Unlike Snap Map, Instagram Map updates location only when the app is open or running in the background, without providing continuous real-time tracking.

Users can access the Map by going to their direct messages and selecting the Map option, where they can control who sees their location, choosing between Friends, Close Friends, selected users, or no one. Even if location sharing is turned off, users will still see the locations of others who share with them.

Instagram Map shows friends’ shared locations and nearby Stories or Reels tagged with locations, allowing users to discover events or places through their network.

Additionally, users can post short, temporary messages called Notes, which appear on the map when shared with a location. The feature encourages cautious consideration about sharing location tags in posts, especially when still at the tagged place.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Beijing 2025 Robot Conference highlights China’s humanoid robotics growth

The 2025 World Robot Conference in Beijing has drawn significant attention to China’s growing humanoid robotics industry. With over 60 humanoid robots on display, the event attracted investors and tech enthusiasts alike, generating a surge in stock prices for companies such as Unitree Robotics.

The conference showcased robots performing diverse activities from industrial operations to more human-like tasks, including marathons and kickboxing, highlighting rapid AI advancements.

China’s Ministry of Industry and Information Technology has supported the sector strongly, aiming for mass production and widespread adoption by 2027. The market, valued at $2.24 billion in 2024, is expected to grow to $41 billion by 2032, reflecting a compound annual growth rate of nearly 44%.

New manufacturing facilities and advances such as carbon fibre materials are boosting the durability and agility of these robots, while companies are focusing on AI integration and teamwork capabilities.

Despite the promising outlook, challenges like high costs, AI learning complexities, and potential overvaluation remain. Experts acknowledge China is closing the gap in humanoid robotics innovation, though technical hurdles persist.

The event underscores the significant role humanoid robots could play in reshaping industries and everyday life, supported by both state initiatives and private investment.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

UAE Ministry of Interior uses AI and modern laws to fight crime

The UAE Ministry of Interior states that AI, surveillance, and modern laws are key to fighting crime. Offences are economic, traditional, or cyber, with data tools and legal updates improving investigations. Cybercrime is on the rise as digital technology expands.

Current measures include AI monitoring, intelligent surveillance, and new laws. Economic crimes like fraud and tax evasion are addressed through analytics and banking cooperation. Cross-border cases and digital evidence tampering continue to be significant challenges.

Traditional crimes, such as theft and assault, are addressed through cameras, patrols, and awareness drives. Some offences persist in remote or crowded areas. Technology and global cooperation have improved results in several categories.

UAE officials warn that AI and the internet of Things will lead to more sophisticated cyberattacks. Future risks include evolving criminal tactics, privacy threats, skills shortages, and balancing security and individual rights.

Opportunities include AI-powered security, stronger global ties, and better cybersecurity. Dubai Police have launched a bilingual platform to educate the public, viewing awareness as the first defence against online threats.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

NASA and Google develop AI doctor for Mars missions

NASA and Google are collaborating on an AI-powered medical assistant designed for long-distance space travel, particularly missions to Mars.

The Crew Medical Officer Digital Assistant (CMO-DA) tool uses Google Cloud’s Vertex AI platform and a mix of open-source large language models to provide autonomous medical advice and diagnostics.

The capability is crucial due to the significant communication delays between Mars and Earth, which can reach up to 223 minutes one way, making real-time medical consultation impossible.

The CMO-DA aims to support astronauts by diagnosing conditions and offering treatment recommendations independently, without Earth-based input. Future versions will incorporate ultrasound imaging and biometric data to enhance diagnostic accuracy.

Tested against common medical issues by a panel that included a doctor-astronaut, the AI demonstrated accuracy rates ranging from 74% to 88% for various ailments.

However, NASA acknowledges challenges such as building trust in AI decisions and limited data on health effects from spaceflight and partial gravity environments.

The project forms part of NASA’s Artemis programme, focusing on Moon exploration and preparing for human missions to Mars, emphasising the importance of autonomous healthcare tools for deep space missions.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Nvidia and AMD to pay 15% share of China AI chip revenue to secure US export licences

Nvidia and Advanced Micro Devices have agreed to hand 15% of their Chinese AI chip sales revenue to the US government in return for export licences.

The arrangement, covering Nvidia’s H20 accelerator and AMD’s MI308 model, is considered unusual and could prove contentious for both companies and Beijing.

The deal reflects Washington’s willingness to link trade concessions to financial payments, but analysts note there is little precedent for such a targeted export levy.

Critics warn the move could undermine the national security rationale for export controls, making it harder to convince allies to adopt similar measures. Beijing, meanwhile, has voiced security concerns over the H20 chip’s performance and alleged vulnerabilities.

Industry observers suggest the payment requirement could discourage further expansion by US chipmakers in China, the world’s largest semiconductor importer, and give local producers an advantage in building domestic capacity.

Chinese firms such as Huawei are already increasing market share amid tighter restrictions on US technology.

The potential sums involved are significant. Before restrictions were imposed, Nvidia had generated over $7 billion in H20 sales to China in a single quarter. In comparison, AMD could earn up to $5 billion annually if full access to the market resumed.

However, uncertainties over demand and regulatory conditions remain.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Kiwi.com eyes agentic AI future with new booking technology

Kiwi.com has unveiled an AI-powered system that enables direct airline bookings, partnering with AIpic to launch the industry’s first Model Context Protocol (MCP) server. However, this technology links flight inventory directly with major AI platforms.

MCP is an open standard likened to a ‘USB-C for AI’. It lets large language models access real-time services beyond their pre-trained data. The access enables AI agents to search and book flights on a user’s behalf.

Kiwi.com says the technology positions it to capture growing demand, as consumers increasingly use AI platforms like ChatGPT, Claude, and Microsoft Copilot to plan travel. Experts anticipate that agentic AI systems will become the dominant interface for online services.

With MCP, users can request flights in natural language, specifying dates, destinations, passenger numbers, and cabin preferences. The AI agent accesses Kiwi.com’s inventory, returning curated results in the user’s preferred currency and time zone and an instant booking link.

The company considers the integration a new distribution channel and a potential model for other online travel agencies. It adapts to changing search and booking behaviours driven by AI.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Musk ends in-house ‘Dojo’ AI chip programme, shifts focus to external partners

Tesla has reportedly shut down its Dojo supercomputer project following multiple high-profile departures, including that of project head Peter Bannon. CEO Elon Musk ended the AI chip programme, reassigning the remaining staff to other data centre projects.

Dojo aimed to process vehicle data for autonomous driving and reduce Tesla’s reliance on Nvidia and AMD. The project faced delays, with leaders such as Jim Keller, Ganesh Venkataramanan, and Bannon departing before its closure.

About 20 former Dojo employees have joined DensityAI, a stealth startup founded by ex-Tesla staff, which is expected to work on AI chips for robots and data centres. Tesla will now rely more on Nvidia, AMD, and Samsung.

Samsung recently secured a $16.5 billion deal to supply AI chips for Tesla’s self-driving cars, robots, and data centres. Musk said Samsung’s Texas factory will produce Tesla’s AI6 chips, with AI5 chips to be made in 2026.

Musk suggested that combining AI5 and AI6 chips could form a ‘Dojo 3’ system, while Dojo 2 would not launch. The shutdown comes as Tesla restructures, with executive exits, job cuts, and renewed focus on AI integration across Musk’s companies.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

James Cameron warns AI could spark a real-life Terminator apocalypse

James Cameron, the director behind the iconic Terminator franchise, has warned that the real-world use of AI could lead to a catastrophic scenario similar to the series’ apocalyptic Judgement Day.

While Cameron is writing the script for Terminator 7, he has expressed concern that mixing AI with weapons systems, including nuclear defence, poses grave risks.

He explained that the rapid pace of decision-making in such systems might require superintelligent AI to respond quickly. Yet, human error has already brought the world close to disaster in the past.

Cameron also highlighted three major existential threats humanity faces: climate change, nuclear weapons, and superintelligence. He suggested that AI might ultimately offer a solution rather than just a danger, reflecting a nuanced view beyond simple dystopian fears.

His evolving perspective mirrors the Terminator franchise itself, which has long balanced the destructive potential of AI with more hopeful portrayals of technology as a possible saviour.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Diplo chatbot can make mistakes. Consider checking important information.