Enterprises will need to nail and scale complex data management

Zuzanna Stamirowska, CEO and co-founder, Pathway

“2025 will see enterprise AI transition from proof of concept to production, with a sharper focus on ROI. However, challenges like data management, privacy, and updates will intensify. To succeed, organisations must feed enterprise data into LLMs, ensuring models can contextualise and retrieve complex data at speed. Live AI will gain prominence, enabling smarter handling of fresh structured and unstructured data for operational success.”

Ellison Anne Williams, founder and CEO, Enveil  

“While the promise of AI remains substantial, its value is rooted in an organisation’s ability to access and use rich, relevant data for evaluation and training. This core need will be abundantly clear as AI tools move from sandboxes to operational environments where the stakes are much higher.

“For cyber leaders, this will also amplify the need to ensure that data sources can be used in a manner that prioritises both security and privacy. If not, the value business and mission value delivered by AI will be unable to overcome the impact of risks it introduced.”

Andrew Beal, chief architect, Markerstudy

“AI’s explosive growth has elevated data’s importance, but most organisations’ data platforms were built for MI and reporting, not AI. While AI requires raw, unstructured data, traditional platforms aggregate and structure data, losing its unique value.

“GenAI will unlock unstructured data’s potential, turning documents, images, and videos into valuable resources and driving a need for robust data ecosystems.

“In 2025, security, governance, and ethical practices will become critical. Tools like data catalogues and tagging will ensure data is well-managed, while organisations will need frameworks to address privacy, bias, and ethical AI applications.”

The democratisation of data access continues…

Martin Brunthaler, chief technology officer and co-founder, Adverity 

“Data democratisation will reach new heights, fundamentally reshaping how businesses operate. We can expect to see more brands and agencies taking steps to ensure data is easily and quickly available: empowering their teams to take full control of insights without heavily relying on IT or data engineers, so IT can establish secure, scalable systems to maintain organisational standards and focus on other strategic, more valuable initiatives.

“Features like conversational data interfaces, smart recommendations, and customisable micro-apps will become standard, making accessing and acting on data simpler and faster.

“Data democratisation will play a pivotal role in enabling business transformation at scale, helping companies to stay competitive in a rapidly changing world.”

Analytics: Dashboards are dead

Derek Slager, CTO and co-founder, Amperity 

“Dashboards are dead – Generative AI-powered tools offering the ability to answer the questions that matter on-the-fly will be the new surface for analytics and decision making.”

Hybrid models will power Big Data

Enrico Signoretti, VP of product and partnerships, Cubbit

“Unstructured data will grow steadily in 2025, making data management essential for storage and value extraction. Organisations will tackle data growth with CapEx, OpEx, and optimisation, favouring hybrid cloud and STaaS models. Advances in visibility will address challenges like data silos from hybrid and multi-cloud usage, with S3 data lakes emerging as a key solution for improved accessibility and security management.”

Chris Hall, CPO, Precisely

“Disruptive technologies like AI and cloud adoption are driving demand for simplified, trusted data access. To meet this, organisations will converge fragmented data ecosystems and adopt unified data integrity strategies across hybrid environments. Integrated data management reduces costs, ensures governance, and enables data teams with self-service access to power analytics and AI initiatives.”

The post 2025 Informed: Big Data predictions for the year ahead appeared first on TechInformed.

AI-driven robotics will be able to learn on the job

Pascal Brier, chief innovation officer at Capgemini 

“Advancements in AI technology have accelerated the development of next-generation robots, building upon innovations in mechatronics and expanding beyond traditional industrial uses. While robotics used to be dominated by hard-coded, task-specific machines, the development of Gen AI is spurring the development of new products (including humanoid robots and collaborative robots – or cobots) that can adapt to diverse scenarios and learn continuously from their environment. According to the Capgemini Research Institute’s upcoming report, 24% of top executives and 43% of Venture Capitalists see AI-driven automation and robotics as one of the top 3 tech trends in data and AI in 2025. With robots becoming more autonomous and AI taking on complex decision-making roles, the future of work may see a shift in the traditional structure of authority. The rise of AI-powered machines that mimic human behaviours challenges our understanding of leadership, responsibility, and collaboration, ultimately pushing us to reconsider the role of humans.

Why it matters: As Industry 4.0 progresses, AI-powered robots will drive efficiency, flexibility, and innovation, becoming key components of intelligent, connected systems that redefine industrial processes. By 2025, advances in natural language processing and machine vision will further enhance their capabilities, allowing robots in manufacturing, logistics, and agriculture to take on more complex roles within the modern workforce.”

Michel Spruijt, president, Brain Corp International

“Polyfunctional robots – those capable of performing multiple tasks and seamlessly switching between them as required – are set to have a growing impact across many industries in 2025. We’ve already seen significant impact in retail where robots seamlessly transition between tasks such as floor care and inventory management.

Looking ahead, we anticipate polyfunctional robots will become increasingly prevalent in manufacturing, logistics, and retail. These robots will be able to adapt to changing production needs, provide value in increasingly complex warehouse environments, and enhance customer experience in a variety of contexts.

One of the most promising aspects of polyfunctional robots is their ability to learn and improve over time. As they encounter new situations and tasks, they are able to adapt to increase efficiency and overall productivity.

In 2025, we expect polyfunctional robots to play an increasingly important role across many sectors, driving innovation, increasing productivity, and creating new opportunities to future-proof industries.”

Robots will be able to adapt to their human colleagues

Volker Spanier, head of manufacturing solutions, EMEA, Epson

“In 2025, we will see a wider adoption of smart manufacturing, which in the UK will be underscored and stimulated by the Government’s new £16 million investment aimed at accelerating digital transformation among SME manufacturers. Alongside this, the expansion of the UK Government’s Made Smarter Innovation programme in 2024, with an additional £37 million in funding for 2025-2026, aims to promote innovation and digitalisation across the manufacturing sector. This programme includes grants for research and development in automation and robotics, as well as tax incentives for manufacturers investing in AI-driven technologies, further catalysing the shift towards smarter, more efficient manufacturing processes.

The increasing productivity and safety of robotics, alongside easier software, will reduce the need for expensive peripheral equipment or complicated programmes for automating operations. Instead, these robots can be quickly and easily repurposed for various products and processes. This year marks a turning point where robotics, with its precision and speed, becomes a partner to humans, enhancing productivity and strengthening the UK’s manufacturing excellence and superiority.”

Adrian Negoita, CTO, Dexory

“Over the past year, the robotics industry has seen impressive progress, both in terms of technological evolution and increased adoption across other industries. It is now firmly an industry for the future that will drive significant change across multiple industries.

As we look ahead to 2025, we can expect to see further improvements in the autonomy and sensor technology integrated into robotic systems. This will allow robots to better navigate their surroundings and handle more complex, unstructured environments. I foresee a world of increased collaboration between humans and robots – this improved autonomy will be essential to that, as it means robots can smoothly operate in the same spaces as humans without safety concerns. This will be key for industries such as logistics, where companies are looking at deploying robots alongside humans to improve efficiency, freeing employees time to focus on strategic planning and driving growth within these industries.

With the continued development and growth of the robotics industry, it is an opportunity for people to upskill themselves for a field that will provide opportunities for growth and allow them to define and shape the future of robotics.”

The post 2025 Informed: Four robotics predictions for the next year appeared first on TechInformed.

With hybrid models gaining traction, AI integration accelerating, and multi-cloud strategies redefining enterprise IT, cloud technology stands at the forefront of digital transformation.

Industry leaders share their insights and predictions for the trends and challenges that will dominate the cloud ecosystem in the year ahead.

The AI cloud revolution

John Bradshaw, director of cloud computing technology and strategy EMEA, Akamai

“In 2025, AI will become integral to Cloud-driven operations, with more tangible applications than before. Organisations will begin to realise the potential of changing company headcounts by automating repetitive tasks and decision-making processes.

However, many companies may face challenges due to underbudgeting for AI implementation and the substantial data requirements necessary for these systems to function effectively.

As AI moves from buzzword to business imperative, many organisations will find that their Cloud infrastructure (essential as the backbone of AI advancements) is not yet equipped to handle the intensive data and computing workloads that AI demands.” 

Hybrid and multi-cloud strategies

Sergei Serdyuk, VP of product management, NAKIVO

“The adoption of the hybrid cloud model will continue in 2025 due to its flexibility, scalability, and cost-efficiency. The hybrid model has highlighted the need for cost optimisation as data volumes grow, as well as the necessity of robust cloud security.

“New tools like standardised APIs, self-service portals, and cloud storage gateways have made it easier to manage hybrid cloud environments. However, more work is still needed to streamline cloud and on-premises systems integration.

“Nevertheless, new challenges are bound to arise, such as the need for consistent security practices across hybrid cloud infrastructures to manage risks effectively. Additionally, the complexity of managing a hybrid cloud infrastructure will increase data sprawl and waste resources and will require skilled professionals to manage it.”

Steve Ellis, head of Amdocs cloud division, Amdocs 

“In 2025, enterprises that initially made big bets on a single cloud hyperscaler will begin to diversify by introducing secondary providers, adding competition, and unlocking capabilities their primary provider may not offer.

“While the major cloud players still dominate enterprise spend, there will be a noticeable shift toward multi-cloud strategies as businesses seek to complement their existing investments. This trend and the evolution of cloud networks from the enterprise to the application portfolio level will drive greater self-service capabilities for both tech and business teams.

Additionally, as  cloud continues to mature, AI’s regulatory compliance and technical tooling will also improve, enabling public-facing AI interactions to become more commonplace.”

Andrew Beal, chief architect, Markerstudy

“In 2025, the adoption of hybrid and multi-cloud environments will continue to accelerate, driven by organisations seeking to optimise costs and avoid vendor lock-in.

As more providers embrace multi-cloud strategies, we will see significant growth in platforms offering a single pane of glass for platforms such as API management, firewalls, data, and Kubernetes. This move would reduce the overheads associated with managing multiple cloud environments.

To support the drive for standardisation, Kubernetes will become the default hosting platform for many, ensuring software engineers have a consistent development experience, regardless of the cloud provider.

Much of this shift will be driven by platform engineering, as organisations transform their DevOps teams into platform engineering teams that focus on the entire developer ecosystem. This transformation will enhance the efficiency and productivity of development processes across different cloud environments.

Overall, 2025 will be a pivotal year for cloud technology, with advancements in hybrid and multi-cloud environments providing businesses greater flexibility, security, and control over their digital infrastructure.”

Data protection and security in the cloud

Johan Paulsson, CTO, Axis Communications

In previous years, we’ve highlighted how hybrid architectures, those making the best use of edge, cloud, and on-premises technologies, have become the de facto choice for security solutions.

The drivers for the choice of architecture will be unique to every organisation, considering technological, legal, ethical and governance concerns and requirements.

Whether demanded by emerging local and regional regulations or concerns over data control, cost, or energy efficiency, hybrid solutions will continue to offer the greatest flexibility in creating architectures to suit specific organisational needs best and allow a system to be scaled.”

Sergei Serdyuk, VP of product management, NAKIVO

“Data protection will always be a top priority for cloud environments. More businesses are using strong cloud security measures to protect against ransomware attacks, including encryption, access controls, and immutable storage to combat cloud-specific cyber threats.

There’s growing understanding of the shared responsibility model between organisations and cloud service providers (CSPs), though some companies are still catching up. The focus will be on implementing robust backup and recovery solutions, encryption, and access controls to protect critical data across hybrid cloud infrastructures.”

Andrew Beal, chief architect, Markerstudy

“The current geopolitical climate, marked by uncertainty and the potential for a global trade war, will bring cloud sovereignty to the forefront. Discussions around data sovereignty and localised cloud services will become more prevalent, with major hyperscalers already exploring solutions in this space.

This shift will be crucial for businesses looking to ensure compliance with regional regulations, such as the Financial Conduct Authority (FCA)  for UK financial services and Insurance companies and mitigate risks associated with geopolitical tensions.”

The post 2025 Informed: Cloud Predictions appeared first on TechInformed.

Healthcare LLM training and specialised AI tools

Oskar Wenbar, COO and co-founder, Evaro

“2025 will mark the end of ‘AI theatre’ in healthcare. After the initial rush to implement any form of AI, we’ll see a decisive shift toward measurable, clinically validated AI applications. The standout innovation will be ‘predictive healthcare triage’ – AI systems that combine patient data, medical histories, and real-time health metrics to predict and prevent health issues before they become critical.

“Two major trends will emerge: the rise of ‘healthcare data cooperatives’ pooling anonymised patient data to train AI, particularly impacting preventative care. The NHS’s dataset will become one of Britain’s most valuable assets.

“Generic chatbots will be replaced by specialised AI tools focusing on specific medical domains, augmenting clinicians’ capabilities—for example, AI specialising in chronic condition management or medication response patterns.

“Data sovereignty will shape this evolution. Countries with robust healthcare datasets like the UK will gain advantages, sparking new collaborations and data-sharing agreements, especially post-Brexit.”

More efficient clinical trials are on their way

Melvin Lai, senior associate, Silicon Foundry

“Technologies using AI, data analytics, and digital platforms will continue to emerge to streamline clinical trials by improving processes, data quality, and patient engagement.

The life sciences sector has been among the lagging industries in the post-Covid era but is expected to outperform in the coming year as capital market conditions continue to improve and a window for IPOs and M&A exits spurs further positive momentum.”

Drug discovery over drug repurposing

Marina Hickson, managing director, Vivanti

“Start-ups like BenevolentAI and Recursion are making a compelling case for the transformative power of AI in drug discovery, and big pharmaceutical companies are starting to take notice. Over the past five years, nearly all major pharma players have dipped their toes into AI, primarily focusing on drug repurposing. However, there’s a noticeable shift from repurposing to developing new drugs, a trend expected to continue into 2025”

Quantum computing and AI will help accelerate innovation

Monika Rai, partner, EIP

“Quantum Computing to enhance AI in healthcare will likely be a standout technology in 2025, accelerating big dataset evaluation for more precise and efficient solutions.

“For healthcare businesses, this offers opportunities to improve clinical trials, speed up medicine development, and drive innovation. However, it also brings challenges, including access to reliable medical data, smaller markets for personalised treatments, and the need for collaboration within a competitive industry and across governments—all amid geopolitical turmoil.

“Meanwhile, traditional healthcare solutions remain unavailable to much of the global population.”

Financial constraints will see rise in tech adoption

David Joerring, CEO and co-founder, HealthKey

“2025 will see global health systems under intensifying pressure, driving rapid innovation in healthtech. Unsustainable financial models will accelerate the adoption of AI triage systems, virtual-first healthcare, and scalable preventative care. Comprehensive virtual healthcare packages, including GP services and chronic disease management, could become available for as little as £50 per person annually.”

Roger Mazella, senior medical product lead, Qt Group

“The MedTech industry is on the verge of a big boom. Recent downturn in the market, might suggest that the industry is headed for a steep decline but a more accurate way to view this is as a necessary course correction of the pandemic-driven industry bloating that led businesses to overexpansion and missteps in strategy.

“The conditions are ripe for a boom. We’ve got the biggest part of the population – the baby boomer generation – approaching their 60s and 70s. Healthcare is increasingly moving out of the hospital, with more care delivered at home. New technology initiatives are advancing to improve the delivery of healthcare for more people, in more locations, in a manner that is more efficient.”

Femtech advances will make workplaces more inclusive

Caroline MacDonald, Founder & CEO, Oggadoon 

“Healthtech, or specifically, Femtech continues to captivate startups and investors alike, showing remarkable growth potential. People are becoming more proactive about monitoring their health. However, a significant challenge remains: most research still overlooks the complexities of the female body.

“In 2025, there will be increased resources for managing menopause and perimenopause symptoms, offering holistic approaches to improve quality of life. In the workplace, we are witnessing a growing trend towards support systems for navigating menopause, with technology set to play an increasingly important role in this transformation. Also, demand to monitor general menstrual cycles, and the rise of Gen-Z, who need to understand their hormones more than ever. “

Valentina Milanova, founder, Daye

“AI and ML will revolutionise gynae healthcare, enabling more accurate diagnostics, especially for underfunded, under-addressed conditions, and personalised treatment plans. We’ll see AI-powered analysis of mammograms and other imaging becoming standard, significantly improving early detection of breast and ovarian cancers.

“Advanced wearables and femtech devices will expand beyond cycle tracking to provide comprehensive health monitoring. These will offer real-time data on hormonal changes, bone density, and cardiovascular health, enabling proactive interventions for conditions like osteoporosis and heart disease, which disproportionately affect women and AFAB individuals.

“Telemedicine platforms specialising in women’s health will expand, offering services like virtual maternity care, menopause counselling, and mental health support. This will create new business models and improve access to specialised care for women in underserved areas.

“Increased focus on data privacy and security in women’s health apps and devices will likely lead to stricter regulations, potentially reshaping the femtech landscape.”

How will geopolitical events influence healthtech?

Mayra Hurtado, CEO and co-founder, Hormony 

“There are a variety of factors at play for femtech, as women’s health is a highly politicised issue. The news of the US President Elect Trump’s victory, for example, may cause a rise in women’s telehealth services to ensure access to care, particularly in states where women are losing rights when it comes to their healthcare and autonomy over their bodies.

“The same could be said for regions undergoing conflict; where healthcare is scarce there is an increased need for women’s healthcare and paediatrics. At the same time, in the UK for example, where the incumbent government has pledged increased support for startups as well as the launch of the Invest in Women Taskforce, we could see renewed interest in investment in women’s health startups and femtech, which would give the industry increased support and facilitate the development of innovation.”

At-home devices will help advance personalised healthcare

Dr Rui Lopes, medical advisor, OMED Health 

“With the rise of at-home kits and devices to monitor all aspects of our health, 2025 will cement bespoke treatment pathways for consumers and patients. Whether it’s due to long wait times for healthcare treatment, or sheer curiosity, more consumers are turning to health and wellness devices that can be used autonomously, or with private clinician support, to track their health and help diagnose and treat various conditions.

“We’ll keep seeing bespoke pathways spring up for an increasing number of conditions, as brands unlock the technology to make this possible.”

Gareth Jones, business development director, HGS

“The future will see the delivery of care and the interface with patients increasingly devolved away from hospitals and surgeries. The need to reduce incremental costs and hospital admissions, as demand from an ageing population increases, will result in an increase in telehealth, adoption of digital devices at point of care, and the use of wearables.

These changes will see AI and machine learning playing increasingly critical roles, analysing vast amounts of patient data to enable more precise diagnoses, predictive analytics, and personalised treatment plans.”

Read more here: The startups tackling the gender data gap to improve female health

The post 2025 Informed: Healthtech predictions appeared first on TechInformed.

This isn’t just a local issue — data suggests that globally, temperatures reached unprecedented highs this July, marking the hottest month ever recorded.

These rising temperatures are putting pressure on a variety of industries, from energy and transportation to agriculture, and preparing for the summer months is becoming a top priority.

For food retailers, the challenge is particularly unique. As an industry already operating on tight margins, supermarkets must keep perishable produce cool and in optimal conditions even while outside temperatures soar.

At the same time, they need to combat skyrocketing energy costs and prevent equipment breakdowns as cooling systems are forced to work harder in the heat.

To meet these challenges, food retailers are turning to advanced digital technologies. These innovations can improve energy efficiency, reduce the risk of catastrophic machine failures, and ensure optimum conditions for fresh produce, all while minimising food waste.

Fresh, safe, and available

For food retailers, keeping produce fresh, safe, and available for consumers is a top priority. This is especially challenging during heatwaves when warmer conditions increase the risk of spoilage and force refrigeration systems to work overtime to maintain the ideal conditions.

In such scenarios, traditional methods have fallen short, making it clear that precision and control are key.

Digital solutions, such as IoT (Internet of Things), have the potential to overcome these challenges, by providing unparalleled visibility and control over the critical equipment in a supermarket, such as the refrigeration systems and HVAC units.

By mining and monitoring millions of raw, real-time data points from these machines, advanced solutions can then analyse the data, detect inefficiencies, predict potential failures, and make automatic adjustments to ensure optimum conditions for produce.

Take refrigeration, for example. Traditionally, the complex nature of managing vast amounts of refrigeration equipment meant all products were chilled to the lowest temperature required by the most sensitive items, like meat.

This one-size-fits-all approach led to unnecessary energy use and often compromised the quality of less sensitive items. Now, with digital technology that integrates third-party data such as merchandising systems, retailers can tailor refrigeration temperatures to the specific needs of each product type, ensuring maximum freshness and minimising waste.

This level of precise, micro-control over operations not only enhances food quality but also drives significant reductions in food waste. It is this unique ability to combine granular control with macro-level outcomes that provides retailers with a robust and resilient approach to machine management, even during the most challenging conditions like heat waves.

Ensuring energy efficiency

Integrating digital technology into retail infrastructure can significantly reduce unnecessary and costly energy consumption, helping to protect retailers’ bottom lines. During periods of extreme heat, cooling equipment consumes more energy as it works harder to maintain regulated temperatures.

Research from Imperial College, London shows that a 2ºC increase in average summer temperatures will lead to a 6% rise in an estate’s refrigeration energy consumption  over the summer months. This highlights the critical need for energy efficiency in food retail operations.

Significant energy efficiencies can be gained and sustained by deploying IoT software to collect and make sense of hundreds of thousands of data points from machines, controls, and systems across entire retail estates.

Such technology can offer powerful solutions by contextualising this data with other connected systems to gain visibility and precise control over machines, allowing them to optimise the machines, as well as the schedules and processes these systems run.

For example, lighting and HVAC systems can be automatically adjusted based on store hours or weather conditions, significantly reducing energy consumption during off-hours or lower temperatures to limit costs.

The real-time data generated by interconnected assets supports a comprehensive and holistic energy management strategy. During warmer months, when refrigeration units are under increased stress, additional load-shedding capabilities can be implemented. This means energy can be strategically redistributed from less critical units, such as those storing beverages, to crucial units that require more power and lower temperatures, like those preserving fresh produce.

By using these advanced digital strategies, retailers not only optimise energy use but also ensure that their operations remain resilient and efficient, even in the face of rising temperatures.

Mitigating breakdown

Leveraging data and IoT technology provides a significant maintenance advantage, especially during heatwaves when cooling systems must work harder and are more prone to faults and failures. In these high-stress conditions, machines are under greater strain, making the risk of breakdowns and energy inefficiencies higher.

By continuously monitoring cooling assets for incremental changes, IoT solutions can identify and alert when an asset performance deviates from ideal conditions and is demonstrating behaviours of a fault or failure.

When performance issues are identified, advanced solutions can immediately and automatically adjust the system to maintain ideal operating conditions, preventing negative outcomes such as increased energy consumption or spoilage of valuable stock. If these automated adjustments do not – or cannot – resolve the issue, the system will promptly alert an engineer to take further action.

This approach to early fault detection and swift intervention is crucial in preventing catastrophic machine breakdowns. It shifts retailers from a reactive maintenance model to a predictive one, enabling them to address issues before they escalate. By catching problems early, retailers can maintain operational efficiency, reduce the risk of costly downtime, and protect their business from the adverse effects of equipment failure.

Heat waves vs Innovation 

Heat waves pose significant challenges for food retailers, including soaring energy bills, increased risk of equipment breakdowns, and the threat of food spoilage due to compromised conditions. These issues demand robust solutions to protect both business operations and product safety and quality.

However, retailers are rising to the occasion with impressive innovation and resilience and by embracing cutting-edge digital technologies, they are not only addressing the immediate challenges of higher temperatures but also transforming their operations for long-term sustainability. These advancements enable food retailers to improve cost and energy efficiency, enhance operational reliability, and reduce food waste, proving that with the right tools, they can thrive even in the face of extreme heat.

Read more here: Deadline for IoT devices to meet new UK security laws strikes

The post How supermarkets are staying cool in a warmer world appeared first on TechInformed.

However, this was more challenging for certain industries than others, notably those requiring high computing or graphical power and lower latencies as part of their core functions.

Take architecture, for example; much of an architect’s IT system needs to be able to draw and render highly detailed models and blueprints that include a lot of datasets to get projects underway.

Furthermore, in a world where digital twins and virtual modelling play an increasingly important role in construction projects, architects need high computing power and GPU, which aren’t always available unless on-site.

Programmes like Revit, AutoCAD, and MicroStation require powerful tools, and PCs that can run them can cost thousands. Also, downtime can cause severe project delays, introducing substantial extra costs.

Aware of this potential threat, Dave Moyes, who is responsible for information and digital systems at architect SimpsonHaugh, began experimenting with Virtual Desktop Infrastructure in 2018.

Amid a major overhaul of the firm’s desktop solutions, with 60 PCs set to be replaced, Moyes began seeking partners for a Citrix and VDI platform and eventually partnered with Ebb3, a consultancy specialising in digital workspaces.

The evolution of End-User Computing: virtualisation, secure browsers, and Microsoft’s growing influence

Moyes admits that some at the firm were concerned about the impact a VDI system might have on shared projects due to the collaborative nature of architecture projects. This led to what he calls a “small initial deployment.”

Then, like so many businesses that have undergone major digital transformations in recent years, “COVID hit.”

“The world changed,” Moyes tells TI, “and all of a sudden, remote work turned into the best thing since sliced bread.”

A few weeks into the pandemic, Moyes moved all of SimpsonHaugh’s data and compute power from its London office to its main offices in Manchester. That would eventually become the firm’s primary data centre, which had been out of the office for almost two years.

As Moyes discussed the return to the office with the rest of the architect’s senior partners, they realised a VDI could be a much more potent solution.

Laying the foundations

“We always need the highest end-compute and graphics, so we normally opt for the highest-end devices available,” he adds, saying architecture, in particular, requires a lot of models and blueprints that need to be rendered in great detail.

“When we turned to VDI, my remit was that the end user shouldn’t be able to distinguish the difference between using a physical box and a virtual one,” he adds.

By early 2022, the IT division had worked on a proof of concept with the London team. Still, this project encountered obstacles, including several applications that hogged all of the computing resources.

Some commonly used applications were “memory hungry” and would not release the memory immediately once shut down, slowing down other applications.

It was IT consultancy Ebb3 – whose clients include Aston Martin Formula 1 –  which suggested bringing in ControlUP and its DEX platform, which would allow for more detailed monitoring of the VDI environment.

This could identify bottlenecks and implement policies to manage them, such as throttling particularly hungry applications.

“Prior to ControlUP, we would be a lot more reactionary. But this gives us the ability to anticipate almost any problems and keep things running smoothly for the users.

“The data you get from it is invaluable. The most important question for IT managers is why their systems are lagging or failing. Why is something happening, and with this, we get that insight.”

Blueprint for success

Moyes says the switch to VDI is part of a wider overhaul that is in keeping with current trends in architecture. New modelling tools, such as digital twins, have enabled those in construction and design to access “more data than ever before,” he explains.

“It is changing rapidly at the moment. In terms of the quantity and quality of data that is being produced for buildings today compared with ten years ago, it is chalk and cheese.

“Part of this is because we can produce much more information; the technology enables us to do that. But that also allows us to give better design intent and more descriptive and prescriptive information for both the design team and the contractors.”

To put this into context, the firm would create around 500 drawings for a major building project a decade ago, but this has now more than doubled.

Despite this, projects are becoming increasingly tighter financially and in terms of timelines, putting pressure on architects to be quick and agile while maintaining high-quality design proposals.

The overall project is part of a £1.21 million investment into SimpsonHaugh’s digital infrastructure across several years.

Solutions, including Citrix Virtual Apps and Desktop, Citrix NetScaler, VMware vSphere Hypervisor, Dell Servers with vSAN storage, and NVIDIA vGPU, have been installed in addition to Ebb1’s efforts and ControlUP’s DEX.

This year, SimpsonHaugh turned to Nutanix to replace its existing storage setup and increased the size of its backbone from 10GB to 25GB.

Additionally, the firm already sees the benefits of VDI, with an estimated £1.79 million in lost earnings saved, calculated based on the reduction of IT downtime.

Moyes is also looking forward. The next stage will likely involve a solution to bring more compute to the edge of the network. He is also closely watching the opportunities that artificial intelligence can offer.

A privately hosted AI system is a key priority for SimpsonHaugh. Moyes divides the task into three areas: word, image, and generative.

Wish upon a cloud: how a charity’s cloud strategy transformed volunteering

Using AI tools to deal with more mundane tasks, such as taking minutes from meetings, will be covered by word, which still requires human validation.

However, he adds that image AI is “one of the areas making a significant difference” for the architects’ firm. Image generation, especially for concepts without precedent, is a vital component.

AI tools such as Midjourney help create concept art that previously had to be drawn and rendered in a much more time-consuming process.

This AI system has been trained on SimpsonHaugh’s own existing images and data, meaning it generates images that adhere to the architectural firm’s editorial guidelines.

This has also involved significant internal training within an AI policy framework that the firm has already implemented.

“We’ve been doing generative AI for a long time in the guise of computational design and single model parametric design,” explains Moyes.

“Where it is beginning to differentiate now is the fact that genAI is becoming much more aware of rules and dimensions without these being preloaded. So, for example, it will now recognise the travel distance for a fire exit or the necessary tread width for a staircase and do this on its own.”

This, he adds, still requires validation, especially as different countries have different building regulations. “It will probably never replace an architect because you need that validation and human input, but the technology is moving rapidly.”

The post How architects at SimpsonHaugh’s built a virtual IT system appeared first on TechInformed.

We must add that it is not the bargain variety you might find on the discount store’s own shelves, or the branded variety that cyber security vendors dispense at trade shows — although he does admit to liking these — but an external Security Operations Centre that can monitor and respond to security incidents around the clock.

“I’m finally getting my SOC! It was approved last week, so I’m happy to be getting a team for 24/7 coverage from a firm already very knowledgeable about our environment,” he says.

McCourt adds that having “eyes on the glass” 24/7 all year round will ensure that alerts are promptly addressed overnight and that triage is handled.

“It’s about having that staff augmentation and those extra capabilities and 24/7 coverage that I’m excited about. Especially leading up to Christmas when staff are off for a period. I want to take time off, too.”

For McCourt, the SOC means he might also have time to reflect on what has been an eventful year for his organisation.

Parent company Pepco acquired rival retailer Poundshop last year and inherited an e-commerce platform, creating new security challenges.

Last month, the retailer also introduced an online web and mobile loyalty programme, Poundland Perks, which has given the team additional GDPR, data protection, and data security issues to manage.

“These developments mean that we have taken a new step in terms of what we’ve had to manage and maintain, and it’s really putting the focus on the digital side of our business,” he says.

“It’s made us look at where we have visibility from in terms of a basic GDPR perspective and whether we have the processes and procedures in place to accommodate these protections.

“It’s also introduced a lot of new challenges not only from a capability perspective but from a capacity perspective,” he adds.

Structure

Given the organisation’s size, McCourt says he manages a relatively small security team. The London-based Pepco Group encompasses 850 Poundland stores in the UK, along with Dealz stores in Ireland and Poland.

The Pepco brand also has 4,500 value stores in 20+ European countries. Another part of the business is China-based PGS, which helps with supply chain management and product acquisition.

“My role is specifically at Poundland and Dealz, and I have counterparts that work independently at Pepco and PGS. We all help support our parent company,” McCourt explains.

“Compared to other categories of business, we have a relatively small IT team for what we manage and maintain. So really, it’s a prioritisation of products and projects as well as ensuring that we dedicate time to increase the resiliency of our internal processes,” he says.

“It all boils down to that security triad of ‘confidentiality, integrity and availability’ into the digital estate,” he adds.

Hungarian phishing attack

According to McCourt, cyber awareness training has always been central to Poundland’s security posture — even before Pepco disclosed that a phishing attack earlier this year caused one of its European branches to lose €15.5 million (roughly $16.8 million).

Initial reports suggested it was a BEC (Business Email Compromise) attack. In contrast, others claim it was a sophisticated phishing attack, in which scammers impersonated trusted entities to trick people into revealing sensitive information.

Because the incident is still being investigated by the Hungarian police, Interpol, and other law enforcement agencies, McCourt says he can’t provide details.

“I can’t comment on what happened, but I can say what it wasn’t; it wasn’t a BEC attack. But ultimately, there were elements of phishing and communication through nonstandard company protocols, which resulted in quite a loss overall,” he says.

In terms of lessons learned, McCourt says the group has added more processes and procedures around how transactions occur.

The key message he wants to convey to staff is that if something doesn’t feel right, they should try and verify it “because, in this case, as well as the issues caused by the attackers, there was also some breakdown in the process,” he adds.

Verification may even involve picking up one of those old-fashioned devices called ‘telephones’ and checking with a line manager or CFO before making large or unusual transactions.

Poundland has been working with KnowBe4, a security awareness training platform that increases awareness through regular security exercises, for about a year.

Initially, McCourt used the platform for small-scale phishing exercises, but it recently expanded to 1,600 users.

The intention now, he adds, is to roll these regular training exercises out at a group level and add the training to its physical retail store environment, bringing the total number of Pepco users to 10,000.

A coffee with…Erich Kron, security awareness advocate, KnowBe4

According to Javvad Malik, Knowbe4’s lead security awareness advocate, most exercises can be reduced to empowering staff to say no to something and ensuring they verify requests.

“That is probably your biggest defence. And if the phone isn’t the norm, it should be the norm.

“Or, some other process needs to be put in place because while it’s true that AI and deepfakes are on the rise, ultimately, they still target human emotion and trigger someone into doing something within a short time while trying to establish some authority.”

Thinking fast, thinking slow

Malik says that nearly every work-based social engineering attack has three core components. The first is the attacker asserting an authoritative identity. “They will either try to claim to be your boss, a partner, or someone important.”

Then, he says, there is ‘the ask’: “This is often strange or not business as usual, like ‘we’ve just swapped banks’ or something that is not a regular occurrence.”

The third aspect is time pressure: “The tone usually is ‘I’m the CEO, and we need to make this payment to a new partner, and we need this to happen within the next two hours, or we’re going to lose the deal.”

Malik says the only way to avoid falling for such scams — whether it’s an apparent physical likeness of a CEO on a video conference call or an urgent email sent from an official-looking address — is “to slow down, take a deep breath, and reevaluate.”

The security expert cites a book by Nobel Prize-winning psychologist Daniel Kahneman, Thinking Fast and Slow, categorising thinking and decision-making into two systems.

“System 1 is your reactive brain, which is saying, ‘I’m pressured, I’m angry, I’m depressed, so I must react quickly,’ and that’s where most mistakes happen,” Malik explains.

System 2 results from slower, more rational thinking, which tends to occur when one takes a step back or consults with someone else.

“So, when you get an email or a video call asking for something unusual. Take a moment, or better still, pass it on to someone else to have another pair of eyes on it. Or validate it, and your perspective becomes much more rational,” says Malik.

“For instance, if you get a letter through the door from HMRC informing you that you owe tax and need to pay immediately, your heart rate will immediately start to increase.

“But if your neighbour comes around and takes a look, they will be much calmer and can say, ‘Oh, look, this is a scam because of this, this, and this.’ It’s easier if you have someone else to bounce off.”

According to McCourt, Poundland is sending out phishing simulations once every few weeks.

“Sometimes the response is ‘You almost caught me with that one!’ but that’s not what this exercise is about. It’s about understanding the red flags and knowing what to do with them,” he says.

“The chances are the scam is going to more than one staff member, so if they know how to use that phishing alert button or to report it to IT Security, then that puts us in a good position to get those alarm bells going,” he adds.

When asked if Poundland has become a target since a store within its group fell prey to an attack, McCourt said: “We’ve had a couple of similar attacks that we could identify rather quickly, which all had the same M.O.

“We identified them very quickly and nipped them in the bud.”

Regarding other preventive measures Poundland takes, McCourt adds that the company continues to use technologies that identify malicious emails using Microsoft and other third-party tools.

Both Poundland and KnowBe4 (which itself was almost infiltrated by a fake employee/North Korean hacker earlier this year) have been publicly open about the security incidents they have experienced.

Do they think there needs to be more openness and information exchange about such attacks between companies rather than a secretive approach that might stem from concerns over reputational damage or loss of consumer or client trust?

McCourt says it would make his Christmas if firms shared a little more. “I think it’s important, especially for those CIOs and CFOs who do horizon scanning. It’s those individuals who start asking questions that I get involved with and start to answer.

“If there were more details in those reports that said, ‘These were the failings, and this is where people need to enhance their security footprint’, I’d be up for that.

“We all want to learn from each other. I don’t care if I’m talking to B&M or my other competitors because it’s all about keeping people safe. As soon as we lose trust in people, the threat field becomes more valuable for other criminals to get involved.”

The post Getting SOCs for Christmas: Poundland’s cybersecurity journey and lessons learned appeared first on TechInformed.

Recent stats state that AI has the potential to unlock £119bn annually in revenue through driving productivity. However, this potential can only be fully realised if the technology gains acceptance from all levels of an organisation, especially from grassroots employees who are, understandably, concerned about its impact on their jobs.

The pressing question then becomes: How can employers alleviate fears around AI while fostering a culture of trust and innovation? It’s a delicate balancing act and HR is uniquely positioned to lead the way.

Innovation vs employee well-being

When it comes to introducing AI into the workplace, openness and transparency are non-negotiable. Employees are naturally wary of the impact AI might have on their job security, and as HR professionals, we need to front-load communication with honesty. It’s essential that employees see leadership as genuinely considering the broader impact of AI and not just the bottom line.

The role of HR is critical here. We are the conscience of the business, often reminding leadership of the people impact while balancing financial goals. It’s not a matter of asking, “Should we be adopting AI?” but rather, “How do we implement AI in a way that’s beneficial for everyone?” AI should not be framed as a replacement for people but as a tool to enhance their work.

By automating tedious, repetitive tasks, we free up employees to focus on more strategic and creative endeavours. Jobs are evolving, not disappearing, and it’s HR’s role to highlight this evolution. Our message to employees should be clear: AI is here to support you, not to sideline you.

In some cases, AI will inevitably replace certain roles or reduce the need for as many people in highly repetitive areas and it’s HR’s responsibility to communicate this honestly, ensuring employees understand the evolving nature of work. Our message should be clear: AI is here to support growth and efficiency, while reshaping roles, not always replacing the value people bring.

Trust through engagement

National surveys often reveal many employees believe AI will lead to fewer jobs, but two-thirds of employees believe it won’t replace them. This indicates a level of optimism, but there’s still underlying anxiety. Employees worry about using AI in their day-to-day roles, fearing they might be judged for leveraging technology to lighten their workload.

The key to addressing this anxiety is in how HR managers can build trust. By highlighting the benefits AI brings – such as space for increased productivity, creativity, innovation and necessary thinking time – we can shift the conversation from fear to empowerment.

It’s about creating meaningful roles that engage employees and ensure they continue to feel valued. Part of our job is to reskill and retrain the workforce to adapt to AI. This doesn’t mean we take responsibility for every employee’s learning journey, but we must actively understand how AI is affecting job roles and where support is needed.  By doing so, we create a workforce that feels equipped and empowered to work alongside AI, rather than being overwhelmed by it.

Redefining the role of HR

HR’s role is evolving alongside the technology we manage. It’s no longer just about managing people; it’s about managing change. We need to take a proactive stance in helping employees see AI as an ally. This means taking the stigma away from “just sitting and thinking” or using AI for initial ideas. Employees often feel guilty if they’re not constantly busy, but some of the best insights come during those moments of reflection.

AI can help employees carve out time for creativity by automating routine tasks, but HR needs to guide this process carefully. We must avoid the risk of lethargy and complacency. AI should be a tool that sparks innovation, not one that leads to laziness. Encouraging employees to use AI to kickstart projects, but not relying on it for everything, helps maintain a balance between human creativity and technological efficiency. I believe that both HRs and management should share how they integrate AI into their everyday work to alleviate the worry and doubt about using the technology in their own projects.

HR also has a responsibility to create guidelines around the use of AI in the workplace. By understanding AI’s role in enhancing work, we can spot when it’s being used inappropriately and intervene to support employees in using it effectively. Usually, it’s the different tone of voice that stands out in an AI generated piece of work and it’s up to the employee’s line manager or HR manager to offer guidance in how to integrate their own personality and tone into their AI generated projects.  These learnings will ensure that AI is seen as a tool for job enhancement, not job replacement, and protects the integrity of the work being done.

Steps for embracing AI

One practical step HR can take is to embed discussions about AI into the employee experience from the start, including during recruitment and appraisal processes. By making AI part of the conversation from day one, we can remove the mystery and anxiety around it. This transparency ensures employees know what is expected of them and how AI will play a role in their day-to-day responsibilities.

The role of HR is to ensure that AI is used responsibly, both by employees and leadership. It’s a collaborative effort, one that involves every stakeholder in the organisation. Empowering the workforce, regulating AI’s use and fostering an environment where AI is viewed as a tool for growth rather than a threat are key to a successful AI integration.

From HR to the future

HR professionals themselves need to be prepared for the shifts AI is bringing. Job design is changing, and we must adapt by incorporating AI-related skills into job descriptions and creating new roles that align with this new era. By staying ahead of the curve and understanding AI’s potential HR can lead the organisation through this transition smoothly.

AI is here to stay and its impact on the workplace will only grow. However, by approaching its implementation with openness, transparency and a focus on employee well-being HR can ensure that AI becomes a tool for empowerment rather than a source of fear.

Our role is to bridge the gap between leadership’s excitement and employees’ concerns while fostering a culture where AI enhances, rather than diminishes, the human element of work.

The post The HR’s guide to supporting employee use of AI appeared first on TechInformed.

At a glance…

• • ✦ Exploding pagers threatened a new era of warfare
• ✦ Social media giants were brought down a peg by law enforcement and legislation
• ✦ That whole Crowdstrike situation…

July: Crowdstrike.

A Crowdstrike outage took down millions of Windows computers

The biggest story of July (and perhaps all of 2024) was the massive IT outage that took down millions of Windows devices thanks to a fudged update by security vendor Crowdstrike.

The outage, caused by an update to Crowdstrike’s Falcon sensor, affected dozens of banks, supermarkets, broadcasters, stock exchanges, airports, and train lines worldwide — leaving Windows users stuck on the so-called ‘blue screen of death’.

More than 8.5 million Microsoft Windows computers crashed, resulting in Microsoft and CrowdStrike issuing grovelling apologies.

Google introduced rules to fight election deepfakes

In a year of elections, including the UK general election (more on that shortly) and the US presidential election, Google introduced new rules in an effort to combat deepfaked election ads.

The Labour party was elected in the UK

July saw the UK elect a new government for the first time. Keir Starmer’s Labour Party won a landslide victory, promising numerous tech developments.

The party’s election manifesto promised a 10-year digital infrastructure plan, national 5G coverage by 2030, a tax overhaul to encourage entrepreneurship and growth, and investment in and regulation of AI.

The conservative think tank Heritage Foundation was hit with a cyber-attack

There was also a cyber-attack on the Republican Party thinktank The Heritage Foundation, reportedly carried out by hacktivist group SiegedSec, which immediately disbanded after its actions drew the attention of the FBI.

Alphabet axed a planned $23bn Wiz acquisition

In M&A news, Google-parent Alphabet bit and agreed to buy cybersecurity firm Wiz for around $23 billion in what would have been the tech giant’s biggest acquisition of all time.

Within days of the announcement of the negotiations, the proposed deal collapsed as internal opposition at Wiz reportedly grew stronger. Reports claimed the boards at both firms had concerns about the deal, including whether it would clear potential regulatory hurdles.

August: Social Media Giants Brought Down to Size

Pavel Durov, the founder of Telegram, was arrested in France

In perhaps the most shocking development in August, the authorities arrested Telegram founder Pavel Durov as he entered France concerning an investigation into online child sexual abuse, drug sales, fraud, and other criminal activity carried out on the platform.

The prosecutors stated that Durov was being held in custody as part of a cyber-crime investigation assessing twelve offences linked to organised crime.

Telegram said, “It is absurd to claim that a platform or its owner is responsible for abuse of that platform.” However, Durov has remained in France ever since, unable to travel.

A “DDoS” attack delayed Musk’s X-clusive Trump interview

Elon Musk made headlines again. The Tesla boss had openly announced his support of Donald Trump in his race to regain the US Presidency. Still, the two men sat down for an exclusive interview to be broadcast on Musk’s X platform (formerly Twitter).

The stream was delayed due to technical issues, and thousands of users complained they could not access the interview. Musk later blamed a DDoS attack for the glitch.

After 40 minutes, the interview went live again on Spaces, and according to Musk’s figures, more than 16.4 million people listened to the conversation within the first 12 hours.

The CAA created a flightpath for drone deliveries

The UK Civil Aviation Authority approved Amazon’s drone delivery service, paving the way for the future of autonomous logistics.

Amazon had already operated drone deliveries in several US states, and the bots do not require human control.

The retail giant was one of six firms the CAA greenlit. The project aims to gather critical data on drone safety and performance, which will inform the development of comprehensive regulations for wider drone operations.

The EU AI Act came into effect

The European Union’s AI Act came into effect in August, enforcing the world’s first comprehensive AI rulebook after four years of deliberation.

The legislation, approved in March, influences businesses within the European Union and has substantial ramifications for technology giants in the US and UK.

Like the GDPR, the AIA’s extraterritorial reach mandates compliance for any AI systems marketed or intended for use within the EU, regardless of origin.

As companies worldwide adjust to the new regulatory landscape, understanding the challenges and opportunities presented by the AIA will be crucial for future-proofing operations and maintaining competitive advantage.

September: Exploding Pagers & Landmark Legislation

Exploding pagers left dozens dead in Lebanon

One of the most shocking stories in September was when news broke about a spate of pagers, walkie-talkies, and other electronic devices exploding across Lebanon, killing dozens and injuring thousands of people.

The ensuing panic reportedly caused local residents to tear batteries from their smartphones, afraid that any electronic device could be affected.

Details emerged that the explosions were part of a coordinated attack allegedly carried out by Israel’s IDF, who were targeting leaders of the terrorist group Hezbollah.

The attacks came in two waves and raised questions about whether we were about to see a new form of warfare involving electronic devices.

The California governor blocked the AI Safety Bill

California Governor Gavin Newsom intervened to veto the controversial AI Safety Bill, claiming the legislation could stifle innovation.

The proposed bill, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047), would have targeted companies developing generative AI. This type of AI can respond to prompts with fully formed text, images, or audio and perform repetitive tasks with minimal intervention.

As one of the first pieces of law focusing on AI safety in the US, it would have required the most advanced AI models to undergo safety testing.

Without the legislation, Senator Scott Wiener, who wrote the bill, said companies could continue developing an “extremely powerful technology” without government oversight.

He claimed Newsom’s block was “a setback for everyone who believes in oversight of massive corporations,” adding that “we are all less safe as a result.”

The EU issued landmark rulings against Big Tech

The European Court of Justice announced dual landmark rulings against Apple and Google in cases around corporate tax avoidance and market dominance abuse.

The rulings, led by European Union antitrust chief Margrethe Vestager, saw Apple ordered to pay Ireland €13bn ($14.4bn) in back taxes, while Google was fined €2.4bn ($2.7bn) for antitrust violations.

The cases were undoubtedly closely observed across the EU as a significant moment for Big Tech’s European tax affairs — especially as the EU’s investigations between companies and member states have faced setbacks.

Just last year, Amazon successfully defended its tax arrangements in Luxembourg in a court battle, and the Commission similarly lost a case involving the Netherlands’ tax treatment of Starbucks, though it chose not to appeal.

AI went Agentic with the Salesforce “Agentforce” launch

As AI continued to develop, Salesforce showed us where the technology is headed as it unwrapped its Agentic AI solution, Agentforce, at a packed Dreamforce 2024 in San Francisco.

Marc Benioff, CEO of the world’s second-largest software vendor, attacked rivals focused on developing copilot and co-owned AI solutions with enterprises. He said businesses wanted out-of-the-box AI tools, and this, he added, was where Agentic AI would play a key role.

Agentforce is a suite of autonomous AI agents that Salesforce claims will “augment employees” by handling service, sales, marketing, and commerce tasks.

Continue reading — Q4 Wrapped Up: The Top Tech Stories of October to December

The post Q3 Wrapped Up: the top tech stories of July to September 2024 appeared first on TechInformed.

Her core research interest is the development of Neuromorphic Computing – a nascent technology inspired by the efficiency of the human brain, with potential applications in AI, autonomous vehicles, healthcare, and space exploration.

Over the course of our chat Majumdar claimed that this next generation of intelligent and adaptive computing and sensing also offers significant energy savings and more effective real-time decision-making, and that the smartest, most efficient LLMs are to be found in nature rather than code.

Majumdar also talks about the joys of being an international scientist and the benefits of the Finnish system, which has enabled her to conduct academic research while raising a young family.

When did you land in Finland?

My husband got a post-doctoral position in Turku when my eldest son was a few months old. We lived there for about nine years, so it will always hold a special place in my heart.

In the beginning I thought ‘let’s take some time off’ but it was quite isolating being on my own all day, so I started looking at options and through the university there I managed to get a scholarship sponsored by a private foundation – which funded my first research project.

What do you like about researching and working in Finland?

The independence you are given worked well for me as a young mother. No one interferes in anything, it’s a very trust-based system. You know your responsibilities and you know what needs to be done. When my son was sick, I was not in the lab, but when he was fine, I worked during my weekend. That kind of flexibility was helpful, and the supervisors were also supportive.

Another thing is that the childcare support is incredibly good. When I dropped my son off at daycare, I didn’t even worry about what he was doing. That’s something I think in those initial years, helped a lot.

What is neuromorphic computing?

It is about making computing more human-centric and studying also how the human brain processes data.

The human brain works completely differently to a computer’s and this requires a complete paradigm shift. The way current computers run means that the AI algorithms are very power hungry. And that’s because these computers were not designed for this kind of task. AI depends on a huge amount of data – billions of data sets – to make it understand what the correct answer or action is.

But humans, and even the smallest of insects, are good at certain things without any specific training. Take one small insect, the navigation path, or the collision avoidance path they have, it’s hard to replicate that in drones or in current machines.

It sounds like our brains are a lot more efficient that ChatGPT!

The human brain works in spikes whenever an event occurs, whereas computers transmit stuff all the time to the cloud which is a lot less efficient and more energy intensive.

So, a 24/7 surveillance camera. Or a health monitor. When there is nothing happening, a computer will continuously transmit data to cloud just not knowing what to look out for because there’s no ‘local intelligence.’

To imitate biological intelligence, the chips need to be more localised – you need to put them closer to the sensor so that it only communicates the relevant data to cloud, allowing it to work more efficiently.

With arrhythmia detection, for example, personalised trained chips can do 24×7 heartbeat monitoring of patients and can do anomaly detection at a much lower power and communication bandwidth cost than the current systems.

Besides healthcare and security surveillance, are there other applications?

Autonomous cars are another. Anywhere where you need chips to integrate and process data from multiple sensory sources simultaneously. Neuromorphic systems can adapt and “learn” from scenarios, improving their functionality over time.

For instance, if a car’s sensor when it is foggy or there is low visibility, the system can compensate by relying on data from other sensors, through associative learning, to make a safer judgement.

In space exploration, its use is also being actively being discussed. Space missions often have limited power sources, such as solar panels. Neuromorphic chips consume far less power than traditional processors, making them ideal for energy-constrained missions.

So, when Mars is at its farthest distance from Earth it takes more than 50 minutes to communicate in one direction so a simple instruction could take two hours. Chips that enable localised, event-based decision-making without the need for constant instructions from Earth, could be critical for autonomous operations in space.

You referred to a 2019 report that said training a single AI model takes up the same amount of carbon as five cars during their lifetime. What does this mean for the future of LLM development?

And that report was published in 2019 – the power demand from LLMs like GPT-4 are increasing exponentially.

The problem is that so far, we can actually keep up with these energy demands, but what will be challenging is when the hardware cannot keep up – in five-or ten-year’s time.

Moore’s Law – which states that the number of transistors on a chip doubles every two years while costs halve – is widely considered to be nearing its practical limits due to physical constraints in semiconductor fabrication. We simply cannot get these chips any smaller or squeeze more onto them.

The artificial intelligence we are seeing today is coming from the software side but if the hardware cannot give enough support to the software when AI workload magnifies enormously, then things will go very wrong.

How much energy would we save using neuromorphic computing?

Individual components wise: they can save a million-fold energy compared to current devices. But then we are still not able to switch fully to new technology. We are currently working towards a hybrid model.

You say there’s a need for innovation not only from device level but also in architecture for hardware support…

Algorithms need to understand the new devices because the ones we are using are based on old hardware. So, if the new algorithm can exploit the full potential of the new hardware, the system could become more efficient.

Are you investigating partnerships in this area?

Yes, in the US we have projects with the University of Massachusetts (UMAS) where they are doing pressure sensor-based computing. There are also other European projects. But hardware fabrication is expensive – you can spend billions on chip design and development – it’s very resource hungry.

Given how expensive it is, shouldn’t enterprises be more involved?

The big players are there- IBM, Intel, Samsung – they are running activities in this direction, but it is very good to have collaboration between industry and university together. It is not a mature technology yet, but it is coming….

What do you enjoy most about your job?

The best thing is nothing is repetitive, it’s something new all the time and it is always something challenging that you need to solve a problem. Also, getting to know so many very brilliant people and having the opportunity to work with the most brilliant minds. It is the human connection – you get to know so many people, their stories and how and why they chose this field. It is nice.

What do you do to relax, and switch off from work?

I enjoy listening to music. I used to be a singer a long time ago. My mother was a very good singer, and she trained me, mostly in songs from our Calcutta heritage. And when I am stressed or sad I find myself back in those words and connected to them. They are a vastly different collection of songs on many different topics, some of them are on nature, love, spirituality. They have the power to touch you.

How do you take your coffee (or tea)?

I am a very big tea fan. I come from the land of Darjeeling, and it is my favourite morning tea. Luckily in Europe now there are lots of sellers. I make it with loose leaves, the traditional way. No milk, no sugar.

Read more here: Will neuromorphic-controlled robots soon become a reality?

The post A coffee with… Sayani Majumdar, Associate Professor, Tampere University appeared first on TechInformed.