Connect with us

Business & Technology

RWS study finds multilingual AI gaps are narrowing

Published

on


RWS has published results from a TrainAI study on multilingual synthetic data generation by large language models. The research found that performance gaps between major and underrepresented languages have narrowed.

Results can also shift markedly between model releases, with changes in quality and cost-related measures that may alter which system is best suited to a given task.

The findings examine how leading large language models perform when generating synthetic data across multiple languages, an area drawing increased attention as companies deploy AI tools in global content and customer-facing workflows.

One of the clearest trends is improved output in languages that have historically received less support from mainstream models. According to the research, the gap between English and lower-resource languages has narrowed significantly in the latest generation of systems tested.

Among the examples cited, Google’s Gemini Pro recorded quality scores above 4.5 out of 5 in Kinyarwanda, a language in which earlier model versions had struggled to generate coherent text. The study also pointed to broader gains across the market, including improvements in GPT and Claude Sonnet models.

Benchmark Drift

Alongside those gains, researchers highlighted what they called benchmark drift. In practice, that means a newer version of a model can perform worse than an earlier one on some tasks, even as the broader market trend points to improvement.

The latest version of GPT trailed smaller models on several content generation tasks where its predecessor had remained competitive. The study also found that tokenizer efficiency, which affects how much text can be processed for a given cost, varied sharply between successive releases.

That matters for businesses comparing systems for translation, content creation and other multilingual applications, because a model upgrade may change both output quality and economics. A model family that performs well in one release may lose ground in the next, while a rival system may improve in specific languages or workloads.

Public rankings alone are not enough for model selection, particularly for companies operating across multiple markets and language groups. Instead, the research recommends testing models repeatedly against specific use cases as new versions become available.

Enterprise Use

The study’s conclusions are aimed at companies building AI into production workflows rather than conducting one-off experiments. For those users, differences in cultural nuance, text coherence and token efficiency can shape both customer experience and operating cost.

RWS, which provides language, content and AI-related services, framed the findings as a case for closer oversight of model choice and for retaining human review in multilingual deployments. It said high-quality, culturally nuanced data remains central to assessing whether a model is suitable for a specific market or business function.

Vasagi Kothandapani, Chief Executive Officer of TrainAI by RWS, set out that position in a statement accompanying the findings. “This study signals a transformative moment that’s not about replacing human expertise, but about elevating it with the right technology,” Kothandapani said.

“As AI becomes more capable across languages, the need for deep cultural intelligence and human validation is more critical than ever. This is why RWS is guiding enterprises into this new reality by integrating these powerful technologies into content workflows with experts in the loop to ensure accuracy, cultural resonance, and brand consistency on a global scale.”

The research also examined less visible technical measures that can influence adoption decisions. One is tokenizer efficiency, which affects the number of tokens a model uses to represent text and therefore has a direct bearing on usage costs.

Tomáš Burkert, Head of Innovation at TrainAI by RWS, said that issue is often overlooked when businesses compare systems. “A model’s real-world value often comes down to specific, frequently overlooked metrics,” Burkert said.

“Factors like tokenizer efficiency, which can make one model 3.5 times more cost-effective than another in certain languages, are critical. The foundation of a successful AI strategy is a continuous validation process, rooted in high-quality, culturally nuanced AI data, to ensure you’re not just adopting a model, but the optimal model to address your unique enterprise requirements.”

RWS said the overall direction of travel in multilingual AI remains positive, with stronger performance across a wider range of languages than in earlier generations. But the central argument is that improvement is uneven and that model selection should be treated as an ongoing process rather than a one-time decision.

The report recommends independent evaluation with each new release to check whether a model still matches the intended use case, particularly when language coverage, quality thresholds and cost control are all under review.



Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Business & Technology

Sparky Space launches AI platform to bridge learning gap

Published

on


Sparky Space, a Berlin company founded by former German Air Force officer Nils Ristau and technology executive Daniel Schmitz, has launched an AI-based work platform now available worldwide.

The platform targets organisations trying to turn training and knowledge into day-to-day execution, a gap the founders argue persists despite heavy corporate spending on digital transformation, skills development and workplace tools. Sparky Space points to industry research suggesting that only 10% to 20% of learning is applied effectively in daily work.

Ristau said he founded the company after drawing on his military background and his belief that many businesses struggle less with access to information than with putting it into practice. In his view, companies often buy training, frameworks and software that remain detached from operational work.

“Most organisations don’t have a knowledge problem – they have an execution problem,” said Nils Ristau, founder of Sparky Space. “Companies are investing in training, frameworks and tools, but too often these sit in isolation from the reality of day-to-day work.

“What’s missing is the bridge between learning and doing. That’s where performance is either won or lost.”

The platform is designed to bring structured methods and AI guidance into live workflows, rather than offering separate learning modules or retrospective review tools. Teams can use it while working through business problems such as product development, strategic prioritisation and the adoption of generative AI in internal processes.

That approach reflects a wider debate in corporate technology over whether productivity and learning tools should remain separate from operational systems or be integrated into them. Businesses in sectors from manufacturing to professional services are under pressure to show returns on training and transformation budgets as economic conditions remain uncertain and teams are expected to work faster.

Ristau linked the product’s design to the discipline required in military settings.

“In military operations, success depends on disciplined execution under constantly changing conditions,” he said. “You don’t have the luxury of separating learning from action – they have to happen simultaneously.

“Modern organisations face similar complexity. Competitive advantage comes from how well teams perform – not just from what they know.”

Schmitz said the software was built for use during normal working routines rather than as another separate system employees visit occasionally. Many organisations already have enough tools and information, he said, but struggle to make them usable in real situations.

“We wanted to create something that teams would actually use in the flow of work,” said Daniel Schmitz, co-founder of Sparky Space. “There’s no shortage of tools or information in organisations today. The challenge is making them actionable.

“Sparky Space is designed to guide teams through real challenges – whether that’s developing a new product, prioritising strategic initiatives or integrating AI into everyday processes – in a way that is structured, repeatable and measurable.”

Use cases

The platform is intended to support innovation and customer-focused product development, agile project and portfolio management, strategic decision-making, leadership alignment and the use of generative AI in workflows. Cross-functional collaboration and product delivery are also among the areas it targets.

The launch comes as companies continue to test how artificial intelligence can be introduced into routine business operations without adding confusion or duplication. For many employers, the challenge is no longer simply gaining access to AI tools, but embedding them in existing processes while maintaining oversight and consistency.

Ristau said that issue helped shape the product’s approach.

“AI has enormous potential, but without the right ways of working, it risks becoming just another layer of complexity,” he said. “Organisations don’t just need access to AI – they need guidance on how to apply it effectively in real situations.

“That’s why we’ve built Sparky Space to combine human-centred methods with AI support, helping teams not only move faster, but also make better decisions along the way.”

The company is entering a crowded market that includes learning management vendors, workflow software providers and a growing number of AI assistants aimed at workplace use. Its argument is that these categories often leave a gap between knowing what to do and carrying it out consistently across teams.

Berlin remains a significant base for software start-ups serving international business customers, particularly in workflow, automation and applied AI. Sparky Space is seeking to tap that market with a product that links management methods with operational use across teams.

Ristau said the companies that succeed will not be those with the most information, but those that can consistently turn that information into action.



Source link

Continue Reading

Business & Technology

BackLite UK launches Knightsbridge digital ad site

Published

on


BackLite UK has launched a new digital out-of-home advertising installation at the Piccadilly Underpass in Knightsbridge as part of its Landmark Series.

The installation, known as The Knightsbridge, was developed with asset owner Wildstone. It features two 2.8m x 14.4m screens on the eastbound and westbound approaches near Hyde Park Corner.

BackLite UK says the location reaches traffic moving through Knightsbridge, Belgravia and Mayfair, with the displays visible to motorists and pedestrians travelling between some of central London’s best-known retail and leisure districts. The two screens generate more than 3 million fortnightly impacts, according to the company.

The site is close to luxury retail destinations, including Harrods and Harvey Nichols. Advertisers already appearing on the screens include Burberry, Club Med and Franck Muller.

London Refurbishment

The launch forms part of a wider refurbishment programme linked to an agreement between Wildstone and Multiply Media Group, the Abu Dhabi-based parent of BackLite UK, covering more than 10 London sites. Multiply Media Group entered the UK out-of-home market through its partnership with Wildstone.

Wildstone owns the Knightsbridge assets, while BackLite UK handles media sales. Westminster Council awarded Wildstone a long-term licence for the underpass through a competitive tender process.

BackLite UK has positioned The Knightsbridge within its Landmark Series, a collection aimed at advertisers seeking large-format digital sites in prominent urban locations. The range also includes The Cube @ Flannels, the Shoreditch Stack and the Wandsworth Roundabout.

At the centre of the upgrade is a pair of 1440 x 280-pixel digital screens supplied by Daktronics, a US LED manufacturer. BackLite UK selected Daktronics 10mm Outdoor Blue displays for the project.

The development adds to competition for premium outdoor advertising sites in central London, where supply is limited and large-format digital inventory near affluent shopping districts remains relatively scarce.

Jack Fleming, Head of Sales at BackLite UK, said: “The Knightsbridge is a fantastic addition to our Landmark Series, fitting perfectly within our most prestigious collection. There are very few OOH assets in the vicinity, and certainly none of this size and calibre. We quickly recognised the site was much more than an underpass, and worked closely with Wildstone to develop something that really stands out.”

Wildstone says the site was always intended for advertisers targeting the premium end of the central London market. The company has been expanding its role as an owner of outdoor advertising infrastructure while working with media operators on digitisation and upgrades.

Andrew Foster, Group Partnerships Director at Wildstone, said: “With its location in the heart of prime central London, we always believed this site was ideally suited to the premium end of the market. We were therefore excited to secure BackLite UK as our media partner, as its focus on luxury brand advertisers meant we were fully aligned on what the upgrade should entail. This is one of a number of high-end refurbishments we’re carrying out in partnership with BackLite UK and, given our success here, we’re excited to see what the future holds.”



Source link

Continue Reading

Business & Technology

UK workers spend more time on admin than European peers

Published

on


Ricoh UK has published research showing that UK workers spend more time on administrative tasks than employees in other surveyed European markets. The study found that UK workers spend 31% of their time on non-core admin.

The research covered workers in the UK, France, Spain, Italy, the Netherlands and Germany. Italy recorded the lowest share of time spent on admin at 21%.

The UK ranked ahead of Spain at 29%, Germany at 27%, France at 26%, and the Netherlands at 23%. UK office workers surveyed reported losing 15 hours a week to administrative work, or nearly two working days.

Only 51% of UK office workers surveyed said they spend most of their day on tasks that deliver direct value. The research also pointed to growing pressure on retention, with 15% saying they had considered leaving their organisation because of admin burdens alone.

Workplace Strain

The study suggests the issue is affecting team relationships as well as productivity. In the UK, 19% of workers said admin creates conflict or tension within their team, while 16% said they feel resentful towards colleagues with lighter admin loads.

Just 21% said administrative work is distributed equally, and 16% reported generational tension, with younger colleagues seen as resisting such tasks. Nearly a quarter, or 24%, said admin limits their productivity, while 21% said it leaves them less motivated or disengaged. Another 19% said it stifles their creativity.

Across Europe, 48% of employees surveyed said they were considering a new role within the next 12 months. The UK findings indicate admin is one of several pressures shaping how employees view their working day and their employer.

Ricoh also found a gap between how workers view the problem and how they think managers respond. Only 20% of UK workers said they feel their employer cares about admin overload, while 27% said managers underestimate the time it consumes.

Technology Questions

The research comes as businesses continue to review the role of automation and artificial intelligence in office work. In Ricoh’s survey, 34% of respondents reported feeling anxious about the prospect of being replaced.

Uncertainty over how technology will be introduced can add to frustration in workplaces already dealing with heavy administrative demands. The figures suggest the debate is not only about job redesign, but also about employer communication and staff confidence.

Ed MacArthur, Practise Lead – Process Automation at Ricoh UK, said: “UK organisations have pushed hard on digital transformation over the past decade, but that investment hasn’t always translated into simpler workflows for employees. Instead, many are dealing with layers of systems, reporting requirements and compliance processes that sit alongside their core role. That creates duplication, manual workarounds and a heavier administrative load than in markets where processes are either more standardised or less fragmented.”

He said the issue also reflected a mismatch between employee expectations and workplace systems.

Time Reclaimed

MacArthur added, “There is also a clear expectation gap. UK employees are used to consumer-grade technology in their personal lives and expect the same level of ease and integration at work. When workplace tools fall short or when automation is introduced without being properly embedded into day-to-day workflows, it adds friction. The result is a disconnect between the technology organisations believe they’ve invested in and the reality employees experience.”

The survey also asked workers what lighter admin loads would mean for their working lives. Nearly a third (31%) said they would enjoy their job more if they had more freedom to focus on creative tasks.

Another 30% said they would use the time to recharge, while 28% said they would invest it in learning new skills. These responses suggest employees see administrative work not only as a drain on time, but also as a barrier to development, recovery and engagement.

MacArthur said, “When a large share of the week is taken up by repetitive admin, it quickly drains motivation and limits the time people can spend on meaningful work. As the pressure builds, for some, it becomes a reason to look elsewhere. At the same time, uncertainty around AI is adding to the strain. Without clear communication and investment in skills, technology risks creating anxiety rather than confidence, especially as organisations add too many tools. Retention comes down to whether people feel supported and see a future for themselves in the company.”



Source link

Continue Reading

Trending