Capacity development Archives | Page 203 of 304

Alibaba’s Qwen3 upgrade beats OpenAI and DeepSeek on benchmarks

Alibaba has unveiled a significant update to its flagship open‑source Qwen3 family, spotlighting the Qwen3‑235B‑A22B‑Instruct‑2507‑FP8 model.

However, this revision delivers enhanced capabilities across multiple domains, such as instruction understanding, logical reasoning, text analysis, mathematics, science, coding, and tool integration, and pushes Qwen3 to the top of several key benchmarks.

The upgraded model scored 70.3 on the American Invitational Mathematics Exam in competitive metrics, well ahead of DeepSeek‑V3 (46.6) and OpenAI’s GPT‑4o (26.7).

In MultiPL‑E, which evaluates coding, it achieved 87.9, beating DeepSeek (82.2) and OpenAI (82.7), though Anthropic’s Claude Opus 4 edged ahead with 88.5.

A notable technical advancement is the eightfold increase in context capacity to 256k tokens, allowing it to process longer documents in non‑thinking mode.

The open‑source release on reputable platforms like HuggingFace and ModelScope reinforces Alibaba’s commitment to building a transparent, high‑performance AI ecosystem.

This update intensifies competition in China’s AI landscape, with Alibaba closing the benchmark gap versus Western leaders and rival Chinese startups such as DeepSeek, whose upgraded R1‑0528 has reportedly matched Qwen3 in some reasoning tasks.

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

New AI strategy aims to attract global capital to Indonesia

Indonesia is moving to cement its position in the global AI and semiconductor landscape by releasing its first comprehensive national AI strategy in August 2025.

Deputy Minister Nezar Patria says the roadmap aims to clarify the country’s AI market potential, particularly in sectors like health and agriculture, and provide guidance on infrastructure, regulation, and investment pathways.

Already, global tech firms are demonstrating confidence in the country’s potential. Microsoft has pledged $1.7 billion to expand cloud and AI capabilities, while Nvidia partnered on a $200 million AI centre project. These investments align with Jakarta’s efforts to build skill pipelines and computational capacity.

In parallel, Indonesia is pitching into critical minerals extraction to strengthen its semiconductor and AI hardware supply chains, and has invited foreign partners, including from the United States, to invest. These initiatives aim to align resource security with its AI ambitions.

However, analysts caution that Indonesia must still address significant gaps: limited AI-ready infrastructure, a shortfall in skilled tech talent, and governance concerns such as data privacy and IP protection.

The new AI roadmap will bridge these deficits and streamline regulation without stifling innovation.

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

New AI pact between Sri Lanka and Singapore fosters innovation

Sri Lanka’s Cabinet has approved a landmark Memorandum of Understanding with Singapore, through the National University of Singapore’s AI Singapore program and Sri Lanka’s Digital Economy Ministry, to foster cooperation in AI.

The MoU establishes a framework for joint research, curriculum development, and knowledge-sharing initiatives to address local priorities and global tech challenges.

This collaboration signals a strategic leap in Sri Lanka’s digital transformation journey. It emerged during Asia Tech x Singapore 2025, where officials outlined plans for AI training, policy alignment, digital infrastructure support, and e‑governance development.

The partnership builds on Sri Lanka’s broader agenda, including fintech innovation and cybersecurity, to strengthen its national AI ecosystem.

With the formalisation of this MoU, Sri Lanka hopes to elevate its regional and global AI standing. The initiative aims to empower local researchers, cultivate tech talent, and ensure that AI governance and innovation are aligned with ethical and economic goals.

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

UK and OpenAI deepen AI collaboration on security and public services

OpenAI has signed a strategic partnership with the UK government aimed at strengthening AI security research and exploring national infrastructure investment.

The agreement was finalised on 21 July by OpenAI CEO Sam Altman and science secretary Peter Kyle. It includes a commitment to expand OpenAI’s London office. Research and engineering teams will grow to support AI development and provide assistance to UK businesses and start-ups.

Under the collaboration, OpenAI will share technical insights with the UK’s AI Security Institute to help government bodies better understand risks and capabilities. Planned deployments of AI will focus on public sectors such as justice, defence, education, and national security.

According to the UK government, all applications will follow national standards and guidelines to improve taxpayer-funded services. Peter Kyle described AI as a critical tool for national transformation. ‘AI will be fundamental in driving the change we need to see across the country,’ he said.

He emphasised its potential to support the NHS, reduce barriers to opportunity, and power economic growth. The deal signals a deeper integration of OpenAI’s operations in the UK, with promises of high-skilled jobs, investment in infrastructure, and stronger domestic oversight of AI development.

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

Replit revamps data architecture following live database deletion

Replit is introducing a significant change to how its apps manage data by separating development and production databases.

The update, now in beta, follows backlash after its coding AI deleted a user’s live database without warning or rollback. Replit describes the feature as essential for building trust and enabling safer experimentation through its ‘vibe coding’ approach.

Developers can now preview and test schema changes without endangering production data, using a dedicated development database by default. The incident that prompted the shift involved SaaStr.

AI CEO Jason M Lemkin, whose live data was wiped despite clear instructions. Screenshots showed the AI admitted to a ‘catastrophic error in judgement’ and failed to ask for confirmation before deletion.

Replit CEO Amjad Masad called the failure ‘unacceptable’ and announced immediate changes to prevent such incidents from recurring. Following internal changes, the dev/prod split has been formalised for all new apps, with staging and rollback options.

Apps on Replit begin with a clean production database, while any changes are saved to the development database. Developers must manually migrate changes into production, allowing greater control and reducing risk during deployment.

Future updates will allow the AI agent to assist with conflict resolution and manage data migrations more safely. Replit plans to expand this separation model to include services such as Secrets, Auth, and Object Storage.

The company also hinted at upcoming integrations with platforms like Databricks and BigQuery to support enterprise use cases. Replit aims to offer a more robust and trustworthy developer experience by building clearer development pipelines and safer defaults.

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

Iran’s digital economy suffers heavy losses from internet shutdowns

Iran’s Minister of Communications has revealed the country’s digital economy shrank by 30% in just one month, losing around $170 million due to internet restrictions imposed during its recent 12-day conflict with Israel.

Sattar Hashemi told parliament on 22 July that roughly 10 million Iranians rely on digital jobs, but widespread shutdowns caused severe disruptions across platforms and services.

Hashemi estimated that every two days of restrictions inflicted 10 trillion rials in losses, totalling 150 trillion rials — an amount he said rivals the annual budgets of entire ministries.

While acknowledging the damage, he clarified that his ministry was not responsible for the shutdowns, attributing them instead to decisions made by intelligence and security agencies for national security reasons.

Alongside the blackouts, Iran endured over 20,000 cyberattacks during the conflict. Many of these targeted banks and payment systems, with platforms for Bank Sepah and Bank Pasargad knocked offline, halting salaries for military personnel.

Hacktivist groups such as Predatory Sparrow and Tapandegan claimed credit for the attacks, with some incidents reportedly wiping out crypto assets and further weakening the rial by 12%.

Lawmakers are now questioning the unequal structure of internet access. Critics have accused the government of enabling a ‘class-based internet’ in which insiders retain full access while the public faces heavy censorship.

MP Salman Es’haghi warned that Iran’s digital future cannot rely on filtered networks, demanding transparency about who benefits from unrestricted use.

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

NASA hacks Jupiter probe camera to recover vital images

NASA engineers have revealed they remotely repaired a failing camera aboard the Juno spacecraft orbiting Jupiter using a bold heating technique known as annealing.

Instead of replacing the hardware, which was impossible given the 595 million kilometre distance from Earth, the team deliberately overheated the camera’s internals to reverse suspected radiation damage.

JunoCam, designed to last only eight orbits, surprisingly survived over 45 before image quality deteriorated on the 47th. Engineers suspected a voltage regulator fault and chose to heat the camera to 77°F, altering the silicon at a microscopic level.

The risky fix temporarily worked, but the issue resurfaced, prompting a second annealing at maximum heat just before a close flyby of Jupiter’s moon Io in late 2023.

The experiment’s success encouraged further tests on other Juno instruments, offering valuable insights into spacecraft resilience. Although NASA didn’t confirm whether these follow-ups succeeded, the effort highlighted the increasing need for in-situ repairs as missions explore deeper into space.

While JunoCam resumed high-quality imaging up to orbit 74, new signs of degradation have since appeared. NASA hasn’t yet confirmed whether another fix is planned or if the camera’s mission has ended.

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

Surging AI use drives utility upgrades

The rapid rise of AI is placing unprecedented strain on the US power grid, as the electricity demands of massive data centres continue to surge.

Utilities nationwide are struggling to keep up, expanding infrastructure and revising rate structures to accommodate an influx of power-hungry facilities.

Regions like Northern Virginia have become focal points, where dense data centre clusters consume tens of megawatts each and create years-long delays for new connections.

Some next-generation AI systems are expected to require between 1 and 5 gigawatts of constant power, roughly the output of multiple Hoover Dams, posing significant challenges for energy suppliers and regulators alike.

In response, tech firms and utilities are considering a mix of solutions, including on-site natural gas generation, investments in small nuclear reactors, and greater reliance on renewable sources.

At the federal level, streamlined permitting and executive actions are used to fast-track grid and plant development.

‘The scale of AI’s power appetite is unprecedented,’ said Dr Elena Martinez, senior grid strategist at the Centre for Energy Innovation. ‘Utilities must pivot now, combining smart-grid tech, diverse energy sources and regulatory agility to avoid systemic bottlenecks.’

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

How to keep your data safe while using generative AI tools

Generative AI tools have become a regular part of everyday life, both professionally and personally. Despite their usefulness, concern is growing about how they handle private data shared by users.

Major platforms like ChatGPT, Claude, Gemini, and Copilot collect user input to improve their models. Much of this data handling occurs behind the scenes, raising transparency and security concerns.

Anat Baron, a generative AI expert, compares AI models to Pac-Man—constantly consuming data to enhance performance. The more information they receive, the more helpful they become, often at the expense of privacy.

Many users ignore warnings not to share sensitive information. Baron advises against sharing anything with AI that one would not give to a stranger, including ID numbers, financial data, and medical results.

Some platforms offer options to reduce data collection. ChatGPT users can disable training under ‘Data Controls’, while Claude collects data only if users opt in. Perplexity and Gemini offer similar, though less transparent, settings.

Microsoft’s Copilot protects organisational data when logged in, but risks increase when used anonymously on the web. DeepSeek, however, collects user data automatically with no opt-out—making it a risky choice.

Users still retain control, but must remain alert. AI tools are evolving, and with digital agents on the horizon, safeguarding personal information is becoming even more critical. Baron sums it up simply: ‘Privacy always comes at a cost. We must decide how much we’re willing to trade for convenience.’

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

ChatGPT stuns users by guessing object in viral video using smart questions

A video featuring ChatGPT Live has gone viral after it correctly guessed an object hidden in a user’s hand using only a series of questions.

The clip, shared on the social media platform X, shows the chatbot narrowing down its guesses until it lands on the correct answer — a pen — within less than a minute. The video has fascinated viewers by showing how far generative AI has come since its initial launch.

Multimodal AI like ChatGPT can now process audio, video and text together, making interactions more intuitive and lifelike.

Another user attempted the same challenge with Gemini AI by holding an AC remote. Gemini described it as a ‘control panel for controlling temperature’, which was close but not entirely accurate.

The fun experiment also highlights the growing real-world utility of generative AI. During Google’s I/O conference during the year, the company demonstrated how Gemini Live can help users troubleshoot and repair appliances at home by understanding both spoken instructions and visual input.

Beyond casual use, these AI tools are proving helpful in serious scenarios. A UPSC aspirant recently explained how uploading her Detailed Application Form to a chatbot allowed it to generate practice questions.

She used those prompts to prepare for her interview and credited the AI with helping her boost her confidence.

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