Google has introduced Veo 3 Fast, a speedier version of its AI video-generation tool that promises to cut production time in half.
Now available to Gemini Pro and Flow Pro users, the updated model creates 720p videos more than twice as fast as its predecessor—marking a step forward in scaling Google’s video AI infrastructure.
Gemini Pro subscribers can now generate three Veo 3 Fast videos daily as part of their plan. Meanwhile, Flow Pro users can create videos using 20 credits per clip, significantly reducing costs compared to previous models. Gemini Ultra subscribers enjoy even more generous limits under their premium tier.
The upgrade is more than a performance boost. According to Google’s Josh Woodward, the improved infrastructure also paves the way for smoother playback and better subtitles—enhancements that aim to make video creation more seamless and accessible.
Google also tests voice prompt capabilities, allowing users to express video ideas and watch them materialise on-screen.
Although Veo 3 Fast is currently limited to 720p resolution, it encourages creativity through rapid iteration. Users can experiment with prompts and edits without perfecting their first try.
While the results won’t rival Hollywood, the model opens up new possibilities for businesses, creators, and filmmakers looking to prototype video ideas or produce content without traditional filming quickly.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Meta Platforms has unveiled V-JEPA 2, an open-source AI model designed to help machines understand and interact with the physical world more like humans do.
The technology allows AI agents, including delivery robots and autonomous vehicles, to observe object movement and predict how those objects may behave in response to actions.
The company explained that just as people intuitively understand that a ball tossed into the air will fall due to gravity, AI systems using V-JEPA 2 gain a similar ability to reason about cause and effect in the real world.
Trained using video data, the model recognises patterns in how humans and objects move and interact, helping machines learn to reach, grasp, and reposition items more naturally.
Meta described the tool as a step forward in building AI that can think ahead, plan actions and respond intelligently to dynamic environments. In lab tests, robots powered by V-JEPA 2 performed simple tasks that relied on spatial awareness and object handling.
The company, led by CEO Mark Zuckerberg, is ramping up its AI initiatives to compete with rivals like Microsoft, Google, and OpenAI. By improving machine reasoning through world models such as V-JEPA 2, Meta aims to accelerate its progress toward more advanced AI.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
What began as a company powering 3D games in the 1990s has evolved into the backbone of the global AI revolution. Nvidia, once best known for its Riva TNT2 chips in consumer graphics cards like the Elsa Erazor III, now sits at the centre of scientific computing, defence, and national-scale innovation.
While gaming remains part of its identity—with record revenue of $3.8 billion in Q1 FY2026—it now accounts for less than 9% of Nvidia’s $44.1 billion total revenue. The company’s trajectory reflects its founder Jensen Huang’s ambition to lead beyond the gaming space, targeting AI, supercomputing, and global infrastructure.
Recent announcements reinforce this shift. Huang joined UK Prime Minister Sir Keir Starmer to open London Tech Week, affirming Nvidia’s commitment to launch an AI lab in the UK, as the government commits £1 billion to AI compute by 2030.
Nvidia also revealed its Rubin-Vera superchip will power Germany’s ‘Blue Lion’ supercomputer, and its Grace Hopper platform is at the heart of Jupiter—Europe’s first exascale AI system, located at the Jülich Supercomputing Centre.
Nvidia’s presence now spans continents and disciplines, from powering national research to driving breakthroughs in climate modelling, quantum computing, and structural biology.
‘AI will supercharge scientific discovery and industrial innovation,’ said Huang. And with systems like Jupiter poised to run a quintillion operations per second, the company’s growth story is far from over.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Global R&D spending now exceeds $2 trillion a year, yet many companies still rely on intuition rather than evidence to shape innovation strategies—often at great cost.
TechNext, co-founded by Anuraag Singh and MIT’s Prof. Christopher L. Magee, aims to change that with a newly patented system that delivers data-driven forecasts for technology performance.
Built on large-scale empirical datasets and proprietary algorithms, the system enables organisations to anticipate which technologies are likely to improve most rapidly.
‘R&D has become one of the fastest-growing expenses for companies, yet most decisions still rely on intuition rather than data,’ said Singh. ‘We have been flying blind’
The tool has already drawn attention from major stakeholders, including the United States Air Force, multinational firms, VCs, and think tanks.
By quantifying the future of technologies—from autonomous vehicle perception systems to clean energy infrastructure—TechNext promises to help decision-makers avoid expensive dead ends and focus on long-term winners.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The UK government has unveiled new guidance for schools that promotes the use of AI to reduce teacher workloads and increase face-to-face time with pupils.
The Department for Education (DfE) says AI could take over time-consuming administrative tasks such as lesson planning, report writing, and email drafting—allowing educators to focus more on classroom teaching.
The guidance, aimed at schools and colleges in the UK, highlights how AI can assist with formative assessments like quizzes and low-stakes feedback, while stressing that teachers must verify outputs for accuracy and data safety.
It also recommends using only school-approved tools and limits AI use to tasks that support rather than replace teaching expertise.
Education unions welcomed the move but said investment is needed to make it work. Leaders from the NAHT and ASCL praised AI’s potential to ease pressure on staff and help address recruitment issues, but warned that schools require proper infrastructure and training.
Education Secretary Bridget Phillipson said the plan will free teachers to deliver more personalised support, adding: ‘We’re putting cutting-edge AI tools into the hands of our brilliant teachers to enhance how our children learn and develop.’
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The African School of Internet Governance (AfriSIG) convened in Dar Es Salaam, Tanzania, from 23 to 28 May 2025, bringing together a broad mix of African and international stakeholders for intensive internet, ICT, and data governance training. As a precursor to the African Internet Governance Forum (AfIGF), the school aimed to strengthen civil society, public, and private sector expertise in navigating Africa’s rapidly evolving digital landscape.
Katherine Getao spoke about cybersecurity and #GenevaDialogue at the African School on Internet Governance (AfriSIG), highlighting the importance of multistakeholder cooperation in implementing cyber norms.
Representing Diplo, Dr Katherine Getao delivered a keynote on ‘Cybersecurity and Cybercrime in Africa,’ emphasising the continent’s urgent need to build strong digital defences amid rising cyber threats. While the challenges are pressing, she pointed out that they also open avenues for youth employment and entrepreneurship, especially in the cybersecurity sector.
Dr Getao also stressed the significance of African participation in global policy dialogues, such as the Geneva Dialogue, to ensure the continent’s digital priorities are heard and reflected in international frameworks. Drawing from her experience with the UN Group of Governmental Experts, she advocated for Africa to be more active in shaping responsible state behaviour in cyberspace.
The event’s panel discussions and workshops further explored how African voices can better leverage platforms like the Internet Governance Forum to influence global tech governance. For Diplo and initiatives like the Geneva Dialogue, AfriSIG was a key venue for aligning African digital development with international policy momentum.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
OpenAI has rolled out its most advanced AI model yet, o3-Pro, delivering significant improvements in reasoning and task complexity while introducing steep price reductions.
The model is now available to ChatGPT Pro and Team users, with Enterprise and Education access coming next week. Developers can also access it via OpenAI’s API.
O3-Pro is designed for high-performance use across technology, education, and science sectors. It supports advanced capabilities such as web browsing, code execution, file analysis, and memory retention during conversations.
Despite these upgrades, pricing has been reduced drastically—87% lower than o1-Pro—costing just $20 per million input tokens and $80 per million output tokens. The base o3 model has also seen an 80% price cut.
Evaluators consistently rated o3-Pro higher than previous models for clarity, instruction-following, and accuracy, with standout results in benchmarks like AIME 2024 and GPQA Diamond, where it beat Google’s Gemini 2.5 Pro and Claude 4 Opus, respectively.
Although the model lacks image generation and Canvas support, its reasoning capabilities mark a major step forward in OpenAI’s AI offerings.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
IBM has set out a detailed roadmap to deliver a practical quantum computer by 2029, marking a major milestone in its long-term strategy.
The company plans to build its ‘Starling’ quantum system at a new data centre in Poughkeepsie, New York, targeting around 200 logical qubits—enough to begin outperforming classical computers in specific tasks instead of lagging due to error correction limitations.
Quantum computers rely on qubits to perform complex calculations, but high error rates have held back their potential. IBM shifted its approach in 2019, designing error-correction algorithms based on real, manufacturable chips instead of theoretical models.
The change, as the company says, will significantly reduce the qubits needed to fix errors.
With confidence in its new method, IBM will build a series of quantum systems until 2027, each advancing toward a larger, more capable machine.
Vice President Jay Gambetta stated the key scientific questions have already been resolved, meaning what remains is primarily an engineering challenge instead of a scientific one.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Meta Platforms is set to acquire a 49 percent stake in Scale AI for nearly $15 billion, marking its largest-ever deal.
CEO Mark Zuckerberg sees The agreement as a significant move to accelerate Meta’s push into AI instead of relying solely on in-house development.
Scale AI, founded in 2016, supplies curated training data to major players such as OpenAI, Google, Microsoft and Meta. The company expects to more than double its revenue in 2025 to around $2 billion.
Once the deal is finalised, Scale AI CEO Alexandr Wang is expected to join Meta’s new AI team focused on developing artificial general intelligence (AGI).
The effort aligns with Meta’s broader AI plans, including capital expenditure of up to $65 billion in 2025 to expand its AI infrastructure instead of falling behind rivals in the AI race.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
In a new blog post titled The Gentle Singularity, OpenAI CEO Sam Altman predicted that AI systems capable of producing ‘novel insights’ may arrive as early as 2026.
While Altman’s essay blends optimism with caution, it subtly signals the company’s next central ambition — creating AI that goes beyond repeating existing knowledge and begins generating original ideas instead of mimicking human reasoning.
Altman’s comments echo a broader industry trend. Researchers are already using OpenAI’s recent o3 and o4-mini models to generate new hypotheses. Competitors like Google, Anthropic and FutureHouse are also shifting their focus towards scientific discovery.
Google’s AlphaEvolve has reportedly devised novel solutions to complex maths problems, while FutureHouse claims to have built AI capable of genuine scientific breakthroughs.
Despite the optimism, experts remain sceptical. Critics argue that AI still struggles to ask meaningful questions, a key ingredient for genuine insight.
Former OpenAI researcher Kenneth Stanley, now leading Lila Sciences, says generating creative hypotheses is a more formidable challenge than agentic behaviour. Whether OpenAI achieves the leap remains uncertain, but Altman’s essay may hint at the company’s next bold step.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
