Cybersecurity

In the past decade, cloud computing has become foundational to the way the world does business. It’s given rise to entirely new categories of services, products and entire business models. From global e-commerce to the cloud AI tools used by millions of small businesses, cloud computing makes it all possible.

This is why the most successful businesses look beyond the cloud as simple IT infrastructure and embrace it as a new strategic capability.

It’s easy to make mistakes, though – and in my work with companies of all shapes and sizes, I see it happen all the time.

When it comes to cloud strategy, I see certain mistakes again and again. So here I’ll overview what I believe are the most common pitfalls businesses will stumble into over the next 12 months, and give some tips on avoiding them.

1. Not Approaching Cloud Strategically

Cloud isn’t just data centers or online storage. If it’s treated as a purely technical exercise in centralizing access to tools and data, we’ll miss out on some of the biggest opportunities.

Garage startups have grown into global titans in the cloud era by using it to create and deliver new products and services that wouldn’t have been possible before. From streaming music, movies and games to work and productivity tools delivered via subscription, to virtual communities and worlds with their own economies and currencies. This involves not simply “lifting and shifting” legacy systems onto shiny new servers but rethinking entire business models. Embedding cloud into broader business strategy and understanding how it enables automation and innovation is key to avoiding this trap in 2025.

2. Overlooking Security

A mistake I frequently see companies make is assuming their cloud provider will take care of all their security needs for them. After all, companies like Google, Amazon or Microsoft must be pretty hard to hack, right? Unfortunately, misconfigurations on the client side or falling victim to one of the 60 percent of corporate data breaches committed by insiders can quickly render all of their sophisticated defenses useless. And cloud architecture itself can also create vulnerabilities, as demonstrated by an attack on Microsoft’s Azure service in 2023. After compromising one Azure business account through social engineering or phishing, hackers were able to “sidestep” into adjacent accounts owned by different businesses and plunder their data, too.

Avoiding the risks associated with cloud involves building a “security-first” culture where best practice is embedded across the organization, from training on how to spot phishing attempts to deploying AI counter-measures against infrastructure attacks.

3. Failing To Align Cloud And AI Strategy

Cloud and AI enjoy a symbiotic relationship. Cloud computing gives AI algorithms scalability, deployability, and access to all the data they need. AI helps cloud services run efficiently, optimize the use of power throughout its vast data centers, and personalize the delivery of service to fit customer needs. It’s frustrating to see many businesses still treating them as siloed technologies, limiting the value of both.

Netflix combined AI viewer analytics with cloud streaming, creating a new business model. Amazon did the same with e-commerce, and pharmaceutical giants like Pfizer have built AI-based drug discovery platforms in the cloud. Bringing these groundbreaking technologies together is critical to reaping the benefits of both.

Failing to take this approach can result in fragmented initiatives that don’t effectively capitalize on the potential of cloud or AI, expensive failures and loss of trust in digital transformation.

4. Not Keeping A Lid On AI Costs

Cloud computing isn’t cheap, and this is doubly true if you’re running AI workloads, with the high resource overheads associated with training and inference. Failing to control costs can easily lead to runaway bills, particularly given the 24/7, always-on nature of most AI services. Stability AI was seen as a huge success among AI companies thanks to pioneering image generation with its Stable Diffusion models. However, it ran into publicized financial difficulties due to the huge bills it was running up with cloud providers. Many other businesses risk similar problems on a smaller scale in 2025, particularly those ramping up AI operations without sufficient budgeting in place.

5. Becoming Locked In

According to a Gartner poll, 95% of businesses say they are currently “locked in” to their existing cloud provider, or that changing providers would be challenging.

Becoming so heavily reliant on any provider’s service that your business depends on them is a bad move. With cloud computing, it could limit your ability to adopt new technologies and pivot to new strategies when opportunities appear.

Avoiding becoming locked into a vendor’s pricing, roadmap, and usage policies is a key reason that many businesses choose to adopt multi-cloud strategies. This can provide the agility needed to shift workloads across different providers as strategic requirements change. Embracing open-source technology, modular architecture, abstraction layers, and virtualization can also help avoid damaging vendor lock-in.

This year, cloud computing will be a more powerful tool than ever before when it comes to driving growth and innovation. For many organizations, it will be integral to their success (or failure) with AI. Working to avoid the errors covered here in your business strategy should be a priority for anyone who doesn’t want to turn opportunity into costly failure.

Computers, the internet and digitization have been major driving forces of innovation over the last 50 years, but classical computing architecture has its limits.

Quantum computing is emerging as a solution to the problem of rapidly cranking up the amount of processing power we can throw at cracking particularly tricky conundrums, such as the vastly complex calculations necessary for accurately modeling the effects of medicines on humans, or predicting extreme weather events.

I’m not really here to talk about the technical differences, but just a quick primer, in case you’re not sure what I’m talking about:

While classical computers are built on binary bits that can exist in a state of on or off (one or zero), quantum computers process information as qubits, which can be zero, one or, due to the strange behavior of physics when modeled at the quantum level, both at the same time!

Difficult as this is to understand without a grounding in quantum physics, the end result is that they are capable of vastly more complex calculations than the classical computers – laptops, smartphones, workstations and data centers – we use every day.

Businesses working on tasks that could be accelerated with quantum computers have a huge opportunity in front of them. That means understanding what they’re good for in order to identify potential future use cases. So, let’s take a look.

What Quantum Computers Will Be Better At

Building machines that aren’t fixed to the rigid on/off logic is a big step towards building more accurate models of hugely complex, real, physical systems; the world around us, nature, the cosmos and the human body don’t operate in binary, after all!

This makes quantum computers superior when it comes to tackling problems involving large numbers of variables, like complex optimization problems, or computer cryptography.

These calculations are used in finance to structure investment portfolios and assess insurance risk, in logistics to determine the most efficient delivery routes, and in material science to develop new plastics and alloys.

Making better drug discoveries is also dependent on our ability to model molecules with an increasing level of fidelity. The chemical reactions and biological interactions involved at the molecular level often don’t follow the 1/0 logic.

Artificial intelligence (AI) is set to be the most transformative technology of the century, and many of the calculations used in machine learning and data analytics, such as pattern recognition, could be accelerated with quantum computing.

And another area where it’s already being predicted to have a big impact is cryptography and cybersecurity. The encryption that keeps the world’s private data safe is based on the difficulty of factoring large numbers – a task that takes classical computers an extremely long time to complete. Quantum computers, on the other hand, can crack them almost instantly, leading to fears that some methods of encryption will become obsolete and a rush to develop newer “quantum-safe” cryptography. If your business relies on keeping information secure, this is something you certainly need to be aware of now!

What Classical Computers Will Still Be Better At

Despite all the excitement around quantum computers, it’s likely that for most of us, classical computers will still be a mainstay of our day-to-day lives.

For hosting and managing email servers, running workplace and productivity software, administering databases and networks, classical computers will remain the backbone of IT infrastructure.

The increased power of quantum computing doesn’t create practical benefits when it comes to running these workloads.

Systems we have today, built around structured databases, cloud storage, and storage and retrieval of large datasets, will remain the domain of classical computers.

So, too, will the billions of low-powered commodity devices, such as the processors embedded in our cars, home appliances, civic infrastructure and industrial machinery.

Quantum computers are expensive, fragile and often need to be housed in environments where the temperature is close to absolute zero. So, for anything that involves computation on a device that sits on your desk and in your pocket, they won’t be very practical.

The Coming Quantum Revolution

To sum it up, classical computers will remain the workhorses of day-to-day business technology. Quantum computers, on the other hand, will be highly-specialized tools, designed for solving particular complex problems.

Quantum computers will not replace classical computers; they’ll work together. Thanks to AI and automation, the interface will eventually be invisible to us, with intelligent agents shifting workloads to whichever platform is the most appropriate.

If you do work in a field or industry that’s vulnerable to being disrupted by the arrival of quantum, the transformation is likely to be dramatic.

In financial services, logistics and manufacturing, as with many other industries, competition for efficiency and the cost reduction it creates is cutthroat; milliseconds matter. This means the emergence of new technology like quantum computing creates opportunities for new leaders to emerge and the status quo to be upended.

It’s time to realize that it isn’t only scientists and computer engineers who need to understand what quantum can do. Professionals, business leaders and decision-makers should be getting to grips with it, too, if they want a head-start on the competition.

Credit: Bernard Marr

Generative AI tools like ChatGPT are changing the way people write. This is both exciting and scary for professionals and those who simply enjoy reading and writing.

The concept of AI taking on what is to us a very simple but intrinsically human task like writing does have big implications. How will it affect human creativity? Will all writing start to sound the same?

I don’t think AI is going to replace good human writers. No one wants to read robotic, generic text that reads like algorithms assembled it.

On the other hand, it can be a powerful tool to help writers organize ideas, structure work, critically assess their own output, and ultimately hone their writing skills.

The difference between getting tools like ChatGPT or Gemini to write for you and using them to improve your own writing is all in the prompts (the natural language instructions telling the AI what you want.)

As with any instructions, more detail means more likely to hit the mark. So here are some prompts designed to help writers, editors and anyone who enjoys communicating and expressing their ideas in writing.

These were written for ChatGPT but should work fine with other chatbots like Google Gemini or Claude.

1. Critically Assess Your Writing

One of the most powerful features of ChatGPT (and the other language-based chatbots out there) is its ability to review work and suggest ways it can be improved. You can cut and paste your writing into the chatbot after the prompt or upload it as a file.

Prompt: “Critically assess this text’s clarity, logic, and factual accuracy. Evaluate how successfully it communicates its core message and addresses reader interest. Check for completeness and that no important aspect of the subject has been overlooked, bearing in mind the length and scope of the text. Provide a report giving steps that can be taken to improve it, for example, further research or more data.”

2. When You Just Can’t Remember That Word

Every writer has that moment when the perfect word eludes them. Just replace the stand-in with the word that comes closest in meaning to the one you want to use!

Prompt: “Please study this text and give me 10 options for replacing the word with a better word:”

3. Smash Through Writer’s Block

Another problem that affects every writer at some point. Some writers say the best way to deal with it is just to write through it. This prompt generates writing exercises that might help you do that (they work for me anyway).

Prompt: “Please generate 5 short writing exercises targeted at helping me overcome writer’s block. My goal is to write a (blog post/novel/speech etc). Ideas or subjects I am interested in writing about include . The exercises should all be aimed at inspiring me to work towards my goal.”

4. Proofreading And Editing

Essential steps that are easily rushed or overlooked. Proofreading should obviously never be left entirely to AI. But it can save a lot of time by giving your writing a first read-through and pointing out easy improvements.

Prompt: “Proofread this text, checking for factual errors, grammatical accuracy, spelling and punctuation mistakes, or typographical errors. Then, edit the text for consistency of writing style, clarity of language, readability, and structure. Provide me with a summary of the changes made during these processes, as well as an edited version of the text.”

5. Persuade Your Audience

Writing persuasively is a valuable skill in many aspects of life, from persuading someone to give you a job interview to appealing a parking fine. This prompt analyzes a piece of text to understand how persuasive it is and to provide tips to make it more so:

Prompt: “The purpose of this text is to persuade the reader that <what?>. Please analyze the content, tone and structure to assess how persuasive it is, and provide a list of tips for improving its persuasiveness.”

about:blankClick to accept marketing cookies and enable this content

Building Better Prompts

The term “writing” covers a huge spectrum of human activity and many disciplines, from drafting technical instruction manuals to novels and plays to the copy we see on every piece of product packaging. Whatever flavor of wordsmith you are, though, it’s likely you can find a way to use AI to make you more efficient, productive, thorough and polished.

Will some writers lose their jobs to AIs? Yes, and I believe the evidence suggests some already have. But those who are willing to adapt to new ways of working and adopt new technologies, like generative AI in their toolkit, won’t find themselves short of opportunities.

ChatGPT and other large-language chatbots are great and have revolutionized many aspects of work. However, they have one big drawback: they talk the talk but can’t walk the walk.

The next wave of AI disruption will be different. AI agents are here—built on the same large-language model (LLM) technology as generative AI but capable of carrying out complex, multi-step tasks with little human input.

They do this by interfacing with other systems— examples we’ve seen include using online shopping and e-commerce sites to make orders and using web design tools to create entire websites.

And if they can’t find a tool to complete the task themselves, they can create their own by designing, writing and deploying their own computer code.

As you can imagine, this has huge implications for the way we work. Those who fear humans will be replaced by machines say it will put even more jobs at risk. Those who think it will give us more free time and help solve important problems say it’s a big leap forward.

So, let’s take a look at how AI agents are set to affect our day-to-day jobs and maybe our long-term career prospects, too.

Automating Manual, Routine Tasks

Processing returns, reviewing invoices and posting updates to field workers are just some of the tasks that Microsoft envisions us delegating to AI agents.

While traditional AI tools can tell us what to do, agents will actually do it for us – identifying the tools and data it needs and connecting with third-party services to get it done.

Microsoft is far from alone. Tech giants including OpenAI, Amazon and Google are all bringing platforms to market aimed at integrating agents into working life.

Backing up this plan with billions in investment and huge sales campaigns, they hope we’ll be keen to palm the tedious and mundane elements of our jobs off onto intelligent assistants. This could include things like data entry, reviewing orders and invoices, answering routine support inquiries, or running security procedures.

Better Data-Driven Decision-Making

Understanding how to work with data to make decisions based on reality has become increasingly important for many professionals. AI agents will give them more power to do this than ever before.

When we talk about automated decision-making at work, we mean systems that can raise alerts, spot unexpected happenings, or optimize systems automatically.

This means that as well as the routine work we talked about earlier, involving following sets of rules, AI can take on “thinking” work, like choosing the best course of action.

Here’s an example. A generative AI can analyze your business and current marketing trends, then create a marketing campaign for you, with all the assets you’ll need to market your business, and make decisions about budgeting for SEO, SEM or whatever it thinks you’ll need.

An AI agent will be able to go further in executing the campaign by interfacing with advertising services, social media, and email marketing tools.

(In theory, it won’t stop there either – it will analyze results, follow up with prospects and adjust its strategy to work better next time.)

There is an enormous challenge for organizations here when it comes to instilling trust that augmented decision-making leads to better decisions. This means decisions that are ethically better in terms of their impact on people and society, as well as better for the business.

Working Alongside AI Co-Workers

Whether we are picking and packing in a distribution center, working in the field or behind an office desk, we’ll become increasingly aware of AIs as “entities” -working alongside us, watching what we do and influencing our work.

Think this sounds creepy? You’re not alone. In 2024, the CEO of an AI HR software firm canceled plans to give AI agents employee records, performance targets and training goals, just like human staff, after receiving widespread backlash.

While we might not be ready quite yet to think of AIs as fellow beings, make no mistake – virtual employees are coming. Businesses will be hoping that the creepy feeling wears off when workers find out how useful and empowering they can be.

The scope here is huge. Automated agricultural machinery that sows, harvests and manages crops. Technical support assistants learn more about customer problems with every call and develop fixes and solutions autonomously. Or simply managing all of your schedules, inboxes and priorities in a truly intelligent way.

High-Value Work

So this is what makes it all worthwhile. With AI agents taking care of everything else, we’re free to concentrate on what really gets us out of bed in the morning. The opportunity to do real, meaningful work that makes a positive change.

For teachers, this might mean less time marking papers and setting assignments, and more face-to-face time with pupils. Customer service agents no longer need to spend time answering routine inquiries and can spend time building deeper connections and understanding client requirements. Managers and executives can prioritize long-term strategic planning, team-building, or pursuing new growth opportunities rather than doing routine administrative work.

In other words, AI agents allow us to spend more time on tasks that create real value, which could lead to richer and more rewarding working lives for everyone.

Human Skills More Valuable Than Ever?

With agents increasingly taking on technical tasks, humans with skills that still can’t be replicated by machines become significantly more valuable. It will still be some time before AI replaces the ability of good human leaders to inspire and motivate teams, resolve conflicts, and deliver real long-term strategy. Those with highly-tuned emotional intelligence, human insight and authentic creative abilities will be more sought-after than ever, as their aptitudes and skillsets increasingly become a key differentiator. A quick piece of very easy career advice – nurture those talents.

Understanding AI Ethics Will Be Critical

For employees during the early days of the agentic AI revolution, this will mean going beyond expecting organizational policies to absolve us of accountability. One 2024 report found that just 21% of workplaces have AI usage policies in place. While this will inevitably increase, the potential for harm to be caused inadvertently by AI – and agentic AI in particular – is enormous.

Writers and artists, for example, claim their work is already being “stolen” and used to train AI, without them being compensated. And Fashion retailer Mango received damaging negative publicity due to claims an ad campaign featuring AI-generated models had been devised without concern for the potential implications for jobs. This is only scratching the surface of the potential ethical challenges anyone wanting to integrate agentic AI into their workflows will have to face if they want to protect themselves from potentially very expensive and harmful mistakes.

How Do I Get Ready For All This?

Because of the potential benefits we’ve covered here, ignoring the arrival of AI agents simply won’t be an option for professionals who want to be at the top of their game.

Understanding how it will impact your own working life, career path, and the future of your occupation, profession, or industryis the first step to getting prepared.

After that, you can start looking for opportunities to use the tools and platforms being made available by US and Chinese AI developers to put it to work.

It’s inevitable that AI will lead to job losses, and logic suggests that the more capable AI becomes, the more jobs will be at risk. Those who want to ensure they benefit from the agentic AI revolution rather than be made redundant by it should be taking action on this today.

What if our workplace success metrics are out of date? We’ve long measured productivity in terms of output—tasks completed, hours logged, KPIs hit. But in a modern, knowledge-based economy increasingly shaped by artificial intelligence, these metrics no longer tell the full story. In fact, they often obscure what really drives performance.

Today, the return on investment (ROI) that matters most is not just about doing more, faster. It’s about creating the conditions for people to do their best, most impactful work—because when people are empowered, engaged, and enabled by the right technology, business outcomes follow.

The Cost of Wasted Human Potential

Let’s start with a reality check.

According to Asana, employees still spend 58% of their time on “work about work”—coordination, duplication, admin, and searching for information. That’s time not spent on strategic thinking, innovation, or delivering customer value.

HP’s Work Relationship Index further highlights the risk: only 27% of knowledge workers feel they have a healthy relationship with work. Fragmented systems and outdated workflows don’t just slow people down—they sap energy, engagement, and innovation.

The cost of wasted human potential isn’t a soft issue—it’s a measurable hit to performance, retention, and growth.

And the stakes are rising: according to HP’s research, 85% of connected knowledge workers feel fulfilled at work, compared to just 52% of those who feel disconnected. Fulfilment is directly tied to engagement, collaboration, and effectiveness—which are, in turn, critical drivers of organizational success.

AI Is More Than Automation—It’s an Enabler of Engagement

Too often, the conversation around AI is framed as “AI vs. humans.” It shouldn’t be. The real opportunity lies in AI for humans—technology that amplifies our creativity, sharpens our insights, and accelerates collaboration.

Yes, AI and automation are freeing us from repetitive, low-value tasks like scheduling and data entry. But that’s just the beginning.

More importantly, AI is creating seamless digital experiences that allow people to focus on higher-order work:

• A product manager can simulate new designs in seconds, exploring more ideas faster.
• A researcher can uncover trends across thousands of studies in minutes.
• And at an enterprise level, intelligent systems help shift IT teams from reactive ticket management to proactive innovation strategies—freeing up capacity for digital transformation initiatives.

This is where we see a new kind of ROI emerging: one rooted in employee enablement, speed to insight, and the ability to drive complex, meaningful outcomes at scale.

Focus Is the New Performance Multiplier

One of the most undervalued drivers of productivity and innovation today is focus—the ability for people to be “in the zone” and fully engaged in high-impact work.

Unfortunately, most digital environments pull us in the opposite direction: pings, toggles, app switching, disjointed platforms. All of this creates friction that erodes attention and slows momentum.

AI-enabled technology plays a critical role here, too. It helps remove digital clutter, automate repetitive steps, and give people a smoother, more connected experience. It’s about designing systems that enable clarity and flow, so people can spend more time on the work that drives results—and less time on the work that just fills calendars.

This is a central theme in HP’s vision: removing the noise of work to unlock deeper impact.

The Executive’s Role: From Technology Leader to Experience Architect

This redefinition of ROI brings a critical new leadership imperative.

Today, executives—especially CIOs and CTOs—must think beyond infrastructure. They are becoming strategic architects of employee experience, ensuring that every layer of the digital ecosystem empowers people to connect better, focus faster, and deliver more value.

This means:

• Choosing tools that streamline workflows, not complicate them
• Designing frictionless digital experiences across teams
• Embedding AI into everyday work—not just behind the scenes, but as a visible, empowering tool for employees

When executives lead with this mindset, they’re not just improving IT performance. They are enhancing engagement, retention, speed, and ultimately, driving business growth.

And employees recognize the opportunity: HP’s data shows that 58% of knowledge workers believe AI makes it easier to collaborate—but nearly half still feel unsure how to use it effectively. There’s a leadership opportunity here to guide, educate, and enable.

Rethinking ROI—And Driving Business Outcomes

AI isn’t the end of work—it’s the beginning of more meaningful, higher-value work.

The companies that thrive in this new era won’t be those that automate the fastest. They’ll be the ones that design intelligently—combining AI, human potential, and frictionless digital experiences to unlock full organizational performance.

Because when you empower people to focus, create, and collaborate at their best, you don’t just change what matters for employees—you drive better business outcomes across the board.

📍 To explore more about how HP is helping leaders rethink work and performance, visit the Future of Work page or access the Work Relationship Index findings.

credit: Bernard Marr

The next wave of AI transformation will be driven by agents. This next-generation AI surpasses chatbots like ChatGPT by not simply generating responses but by taking action.

Unlike chatbots, these systems carry out complex tasks with little human involvement. In coming years, we may see them evolving into virtual assistants, teachers, employees, carers and even friends.

Right now, everyone’s waiting for AI agents to have their “ChatGPT moment”—when millions of people across the world suddenly realize it could actually be very useful. Now, in the wake of the market turmoil caused by DeepSeek, a new Chinese contender has emerged.

The developer of Manus AI claims it’s “the first AI agent … a truly autonomous agent that bridges the gap between conception and execution” and “potentially a glimpse into AGI”.

So, what is it, what does it do, what are its political implications in a world where superpowers are racing for AI supremacy? And is it really a step towards the AI holy grail of AGI?

So, What Is Manus?

Manus is currently an invite-only research preview, immediately casting doubt on developer Monica’s claim that it’s the first AI agent. OpenAI’s similar Operator agent model is also a research preview but is at least available to customers – if only for those paying for the $200-per-month ChatGPT Pro tier.

But a developer video has been released showing it performing real-world tasks like screening resumes, researching housing and analyzing stocks.

Its name, derived from the Latin “Mens et Manus” (“mind and hand”), reflects its structure: LLM algorithms as the mind and deterministic algorithms as the hand. Unlike ChatGPT, which only generates responses, Manus can execute actions—integrating with services, processing data, and performing operations like a traditional computer program.

For example, ChatGPT can write Python code, but Manus can autonomously execute, test, and refine it. Instead of summarizing financial reports, it fetches real data and runs its own analysis.

Interestingly, its creators say that coding doesn’t have to be a goal for Manus, which actually thinks of it as a “universal tool for solving problems.” This suggests Manus could autonomously write and execute code to address challenges we’ve never thought to tackle through programming—expanding AI’s problem-solving horizon beyond our current imagination.

One other important point to consider is that although Manus is currently classified as a closed-source, proprietary engine (just like ChatGPT and Operator), it’s built on open-source technology. The LLM “mind” is built on Anthropic’s Claude LLM, as well as Alibaba’s Qwen. The developers have said that they plan to open-source Manus itself – or parts of it – in the near future.

about:blankClick to accept marketing cookies and enable this content

Is This AGI?

No, it doesn’t qualify as AGI – as the developers themselves state in their video, it’s a “glimpse” of AGI – artificial general intelligence.

AGI is best thought of as an eventual goal of developing AI that can learn to carry out just about any task, much like a human can. It uses the tools it’s given or, if it doesn’t have anything suitable, can work out how to find the tools it needs, or even design and use new tools.

It seems best to think of Manus as a collection of tools and models operating under agentic control, rather than one integrated software entity (like ChatGPT4o, for example).

It is, however, a clear sign that AI is evolving as fast, or perhaps faster, as expected. In a little over two years, since ChatGPT was released, we have seen AIs move from reactive, question-and-response systems to agents capable of actively managing digital and real-world tasks.

Geopolitical Implications

In recent years, we have seen AI become a geopolitical battleground, primarily between the U.S. and China.

The leadership of both nations has made it clear they recognize its enormous potential to shape the future due to its impact on everything from medical research to the economy to warfare.

Historically, it’s fair to say, U.S. companies have dominated the market, with OpenAI, Google, Microsoft and so on, leading on research and bringing products to market.

China, where the government plays a more hands-on role, has made significant strides to close the gap, however, first with DeepSeek and now potentially with Manus.

DeepSeek caused a crash in the price of U.S. tech stocks when it appeared that a Chinese model was able to closely match the performance of top-ranked U.S. models, at a fraction of the cost.

If Manus proves to be as useful as its developers predict, it could signal a further shift in AI dominance, potentially prompting Western players to rethink their strategies.

Everyday Agents

So, the race is on to create what I feel will become the equivalent of the computer operating system for the AI era. Essentially, agentic AI that enables anyone to do things with machines that previously would have taken expert-level knowledge and skills.

Because of this, AI and automation will become increasingly ingrained in daily life as we use them to assist in many everyday and workplace tasks.

This, in turn, will lead to a fundamental shift in the nature of the relationship between humans and technology as, for the first time in history, it becomes capable of acting autonomously on our behalf.

Somewhere in your organization, an AI project is dying. Perhaps it’s the recommendation engine that was supposed to boost sales by 30%. Maybe it’s the predictive maintenance system that promised to slash downtime. Or the customer service chatbot that was going to revolutionize response times. The digital dust gathering on these ambitious initiatives represents not just wasted resources but shattered expectations that make future innovation harder to champion.

The Expectation-Reality Gap

Think of AI projects like icebergs. What executives see in vendor presentations and tech magazines is the gleaming tip above water – the finished, polished success stories. What remains hidden is the massive underlying structure of data preparation, infrastructure requirements, talent needs, and organizational change management that makes those successes possible.

This expectation-reality gap is perhaps the most fundamental reason AI projects fail. There’s a persistent mythology that AI is a magical technology you simply “apply” to business problems like a high-tech bandage. The truth is messier and more demanding.

Consider what happened at a global consumer goods company I advised. Their executive team, inspired by presentations showing how AI could optimize supply chains, commissioned a $2.5 million initiative to do exactly that. Twelve months later, they had sophisticated algorithms that were essentially unusable because nobody had addressed the fragmented, inconsistent data across their twenty-seven legacy systems. The AI solution was like buying a Formula 1 car when you only have dirt roads to drive on.

about:blankClick to accept marketing cookies and enable this content

Flying Without Instruments: The Data Dilemma

If there’s one factor that dooms more AI projects than any other, it’s poor data quality and governance. Organizations consistently underestimate both the quantity and quality of data required for AI to function effectively.

The reality is that AI systems are fundamentally data processing engines. Feed them poor data, and you’ll get poor results – a principle computer scientists call “garbage in, garbage out” that has existed since the 1950s but somehow keeps surprising executives.

A healthcare system I worked with wanted to use machine learning to predict patient readmissions. Six months into development, the team discovered that their historical patient records – the data they were using to train the AI – contained significant biases in how various conditions were coded across different facilities. The AI was learning these inconsistencies rather than genuine medical patterns. It’s like trying to teach someone a language using a dictionary where half the definitions are wrong.

Missing The Human Element

Another fatal error is treating AI implementation as purely a technical challenge rather than a socio-technical one that requires human adoption and integration.

I recall a manufacturing firm that spent $1.8 million on an AI system to optimize production planning. The technology worked perfectly in testing, but on the factory floor, supervisors continued using their traditional methods and simply ignored the AI’s recommendations. Why? Because no one had involved them in the development process, explained how the system worked or addressed their legitimate concerns about how it would affect their roles.

AI initiatives don’t fail in isolation; they fail within human systems that are resistant to change. The best technology in the world is worthless if people don’t use it.

The Strategy Disconnect

Many AI projects begin with a critical flaw: they lack clear connections to genuine business problems and strategic objectives. They’re solutions in search of problems rather than the other way around.

I’ve watched organizations launch AI initiatives because competitors were doing so or because the C-suite read about the technology in a business magazine. These projects inevitably fail because they’re not anchored to specific, measurable business outcomes.

Think of it like building a bridge. You wouldn’t begin construction without knowing exactly which riverbanks you’re connecting and why people need to cross. Yet companies routinely embark on AI projects without defining what success looks like or how they’ll measure it.

Talent And Governance Shortfalls

The AI talent gap remains enormous. Data scientists are in short supply, and those with the rare combination of technical expertise and business acumen are as scarce as diamonds in a sandbox.

Beyond talent, many organizations lack proper governance structures for AI initiatives. Who owns the project? Who makes decisions when trade-offs arise between speed, cost, and quality? Without clear accountability and decision frameworks, AI projects drift into ambiguity and eventually failure.

A telecommunications company I worked with had seven different departments independently developing AI solutions with no coordination. This resulted in redundant efforts, incompatible systems, and eventually, multiple project cancellations after millions were spent. It was digital Darwinism at its worst – initiatives competing for resources rather than collaborating toward common goals.

Skipping The Foundation Work

Think of enterprise AI as a house. You can’t build the roof before you’ve laid the foundation and framed the walls. Yet organizations routinely attempt to implement advanced AI capabilities before establishing basic data infrastructure and analytics competencies.

AI isn’t a technological leap; it’s an evolution that builds upon existing capabilities. Companies that succeed with AI typically have already mastered data warehousing, business intelligence, and traditional analytics before venturing into machine learning and other AI technologies.

A retailer I advised wanted to implement personalized, real-time pricing based on AI. But they couldn’t even produce consistent weekly sales reports across their stores. They were attempting to run before they could walk, and predictably, the project collapsed under its ambitions.

The Path Forward: Making AI Projects Succeed

The high failure rate of AI initiatives isn’t inevitable. Organizations that approach AI with appropriate planning, resources, and expectations dramatically improve their odds of success.

Start with problems, not technology. Identify specific business challenges where AI might provide solutions and articulate clear, measurable objectives. This anchors the project in business reality rather than technological possibility.

Invest in data quality and infrastructure before algorithm development. Remember that AI systems are only as good as the data they consume. Create a solid data foundation before attempting to build sophisticated AI capabilities upon it.

Treat AI implementation as organizational change, not just technology deployment. Involve end users early and often, and consider how AI will integrate with existing workflows and human judgment.

Take an incremental approach rather than swinging for the fences. Begin with modest pilot projects that deliver quick wins, build organizational confidence, and provide learning opportunities before scaling.

Establish clear governance, including ownership, decision-making frameworks, and success metrics. Define who has the authority to make critical decisions when (not if) trade-offs become necessary.

Beyond The Hype Cycle

AI isn’t magic – it’s a powerful set of technologies that, when properly implemented, can deliver extraordinary business value. However, that implementation requires rigor, realism, and resources that many organizations underestimate.

The companies that succeed with AI aren’t necessarily those with the biggest budgets or the most advanced technology. They’re the ones that approach AI with clear eyes about what it can and cannot do, build proper foundations before reaching for sophisticated capabilities, and understand that technological change is inevitably also human change.

The graveyard of failed AI projects needn’t grow larger. By learning from these common mistakes, organizations can ensure their AI initiatives deliver on their promise rather than joining the ranks of expensive digital disappointments.

The world has never changed this fast and will never move this slow again. This stark reality, shared by ServiceNow CEO Bill McDermott during our recent conversation, perfectly captures the transformative whirlwind businesses now face with the emergence of agentic AI – autonomous AI systems that can perform tasks without human intervention.

“The AI revolution is faster than all of them,” McDermott told me, comparing it to previous technological shifts. “I think people have underestimated the importance of technology, transparency, metrics, measurement, and the kinds of things that leaders have to put into their companies now to both get people moving again and to help people catch this wave.”

This isn’t hyperbole. We’ve entered what I’ve been calling an era of hyper-innovation, where AI capabilities double every few months rather than years. For company leaders, this acceleration creates both immense opportunity and existential threat.

The Control Tower Imperative

Perhaps the most significant insight from our conversation was McDermott’s emphasis on the need for integration through what he calls an “AI agent control tower.” This recognizes a fundamental problem most organizations face: decades of building siloed systems that can’t effectively communicate.

“These companies are bogged down with these departmental systems that they keep investing in, keep upgrading,” McDermott explained. “They have different release levels. Some are in the cloud; some are on-premise. Someone’s in someone else’s cloud, and the chaos is only growing. And then everybody in the industry is talking to them about agents for their silo, which is only going to exacerbate the complexity.”

This fragmentation problem is particularly acute with AI implementation. Companies rushing to deploy AI agents in disconnected systems risk creating a new layer of incompatible technology, magnifying existing inefficiencies rather than solving them.

The solution? According to McDermott, organizations need “a single platform with one architecture, one data model, and one gorgeous user experience that serves as the control tower for AI business transformation.” This approach enables the integration of multiple AI agents and diverse data sources – from structured database information to unstructured documents on personal devices.

The Human-AI Workforce Evolution

One of the most fascinating aspects of our discussion centered on how AI agents are essentially becoming “tenured employees” – working 24/7, without vacations or healthcare requirements. This shift is fundamentally changing workforce dynamics.

ServiceNow has already realized nearly half a billion dollars in economic value through AI implementations, with agents freeing up 3 million hours of employee capacity. They’ve seen an 85% customer self-service rate through AI agents (not chatbots), reducing issue resolution time by 80%.

“We now have a new workforce in addition to the human workforce,” McDermott noted. At ServiceNow, they’ve estimated they can perform the same work with 10% fewer human employees, but as a growth company, they’re focused instead on deploying AI agents to enable greater growth with greater efficiency.

This doesn’t necessarily mean mass unemployment. As McDermott reflected, “In 1966, Time Magazine basically made the statement that with the evolution of the computer and then thereafter the personal computer, it would be likely that 90 percent of the jobs in corporations would no longer be necessary… Well, millions and millions of jobs later, we know that’s not true.”

Instead, McDermott envisions a renaissance in productivity: “This agentic AI workforce will do the soul-crushing work in every industry, in the tasks and productivity areas that humans don’t want to do anyway.” The future he describes is one of collaboration, where “thinking machines will build better thinking humans and better outcomes for the global economy.”

about:blankClick to accept marketing cookies and enable this content

Escaping The Legacy Trap

AI investment is surging across industries as organizations race to capitalize on its transformative potential. Yet McDermott identified a critical problem: “80 to 90 percent of the budget is going to keeping the trains running on time, as they’ve always have. And in a lot of cases, that also means upgrading the old tech, which is still old, to newer versions of the old tech, which is a major mistake.”

This approach traps investment capital in legacy systems rather than funding AI innovation. His recommendation? “Stop doing what you’ve always done and take a fresh look at the art of the possible. And I would say you should start moving half of your IT discretionary budget to the AI revolution.”

The consequences of maintaining the status quo are severe. “CEOs are really, really upset because they’ve thrown billions of investment at digital transformation only to find out that 85 percent of the digital transformation projects have not delivered a positive ROI for them,” McDermott shared.

Leadership Principles For The AI Era

When I asked McDermott about leadership in this transformative time, his response revealed principles that apply across industries:

1. Take care of customers above all else: “The only thing that really mattered was my customers. They were my only boss. And if they didn’t come back and I didn’t make them happy, then I lose the game to my competition.”
2. Increase your drive during inflection points: “When I see inflection points in markets like the one we’re in right now, my intensity and my drive only increases. Because I know that I’m loving people, and I’m building trust with people by letting them know that they have to accelerate the pace.”
3. Understand the details to dream bigger: “Look into the details because the details will help you form a bigger dream. It’s not just about technology. It’s about what the technology can do for the dream.”
4. Think exponentially: “If you take 30 linear steps, you’ll go 1 to 30. If you take 30 exponential steps, 1, 2, 4, 8, by the time you get to 30, you traveled around the world. You did a billion steps.”
5. Lead from the front lines: “The way I try to do my job is not do my job behind a desk. But be on the front lines with the people really understanding what we got to do to help the customer win in every industry.”

A Decisive Moment

McDermott’s parting wisdom should resonate with every business leader: “This is a do-or-die moment for both individuals and companies to catch this lightning-fast AI revolution.”

The companies that will thrive aren’t merely implementing AI – they’re reimagining their entire organizational structure and technology stack to create integrated systems where AI agents work harmoniously with humans and each other. Those who cling to old ways of working risk being left exponentially behind.

As McDermott so powerfully stated, “If they don’t catch the wave now, there’s a great chance that they could get left behind, and there’s an even greater chance that companies that don’t catch this wave will, in fact, get left behind.”

The future belongs to those brave enough to reverse engineer the world they want and lead their organizations through this transformative moment. The AI revolution waits for no one.

Spotify has a long history of bringing AI-powered features and functionality to users. This has helped it to grow from a startup to a household name in the face of some big competitors like Amazon, Apple and Google.

When news leaked of a new Music Pro tier, a feature reports zeroed in on was the upcoming “AI-powered music tool.”

The Swedish company is open about its commitment to delivering new listening experiences through AI, so it’s natural that there’s speculation this could be a step towards letting users create personalized versions of their favorite songs.

It’s also further evidence that platform owners are increasingly seeing AI as an opportunity to add premium-tier features that can be used to drive higher revenue.

Some background: Artists, record companies and streaming service providers are currently negotiating a major shake-up of the industry with the intention of kickstarting Streaming 2.0.

So, what could this mean for the platform that popularized streaming music and the future of how we listen to music?

Spotify And AI

Spotify is a success story when it comes to engaging users with AI.

From the early days, features like its New Music and Discover playlists have been built around machine-learning algorithms that match listeners with tunes.

Newer features have let users create “blended” playlists of songs that match two users’ tastes and offer suggestions for fleshing out user-generated playlists.

Behind the scenes, it uses AI to automate the tagging of tracks to allow them to be categorized by content or mood.

And AI insights are also used to power the Wrapped personalized annual summaries of what users have been listening to.

Newer initiatives include the AI DJ launched in 2023, which uses AI-generated voice to join tracks like a radio DJ.

And most recently, AI Playlists let music fans use a ChatGPT-like language interface to describe the playlists they want.

I tried this one recently, asking for “background music for reading Lord Of The Rings”, which gave me songs that were mostly from the movie soundtrack, and “Music for a road trip through mountains,” which gave me a stirring mix of opera, heavy metal and video game soundtracks.

This all shows that Spotify has a history of looking beyond simple recommendation engines to create AI tools focused on generating user interaction and engagement. And it doesn’t look like they’re going to stop anytime soon.

Personalized Music Remixes?

We don’t know much about Spotify’s upcoming AI remix tools, but by studying their past form, we can make some educated guesses.

Personalized music remixing has been something promised by advocates for AI in the future of streaming services for a while.

There are some pretty big hurdles, most prominently the question of rights. Are artists and record companies happy with the idea that a million personalized remixes of their treasured IP could suddenly explode into existence?

It’s been speculated that this might be why Spotify isn’t yet ready to publicly announce a launch date for the new premium tier service.

But it’s a pretty enticing idea. Imagine having the tools to change tempos, reimagine songs in different genres or with different instruments, or replace vocals with entirely different ones at your fingertips.

It’s about offering ways for listeners to become active rather than passive consumers of music. Although much of this is speculation at the moment, I think we can expect to see Spotify and other streaming services take this direction as they build new user experiences.

The Premium-Priced Future Of Music Streaming

Spotify serves as a great example of a web and AI-native business with a mature AI strategy, to the point it can create entirely new tools and services for its customers.

This is a significant step ahead of those stuck at the stage of simply adding chatbots to the list of ways of doing things we can already do.

Perhaps the other insight we can take is that AI features are increasingly being used to differentiate between standard and premium subscription tiers.

We’ve seen this already, with Google adding Gemini access to its premium-tier Google One subscriptions and X (Twitter) giving paid-up users unlimited access to Grok.

I think it’s fairly clear this will strategy will grow in popularity as AI service providers look to recoup their investment in training AI models and hiring human talent.

It can be argued that Spotify has already done a lot to change the way we listen to music by popularizing legal streaming services. And it looks like it still has some new tricks up its sleeve.

How will this prepare it for a future where the only certainty is that the way we consume music and other media will continue to evolve? Time will tell.

YUM Brands, the parent company of Taco Bell and operator of 60,000 restaurants worldwide, has unveiled an AI-powered restaurant manager with the aim of bringing agentic AI capabilities to fast food.

Agents represent the latest wave of AI innovation, capable of complex, multi-step tasks with minimal human intervention. They are considered to be the next generation of cutting-edge AI applications, following generative AI chatbots like ChatGPT.

Some believe that they will pave the way for “virtual employees” that will work alongside humans, augmenting our capabilities and managing routine tasks so we can get on with the fun stuff.

And they are also seen as a stepping-stone on the road to artificial general intelligence (AGI) – the “holy grail” of building AI that can do just about anything we can.

The fast-food industry exists in a state of perpetual digital transformation. Now, YUM Brands has demonstrated its plans for the next leap forward with its Taco Bell franchise, and competitors are also eyeing opportunities.

So, let’s take a look at how fast food is cooking on all cylinders with agentic AI.

How Is Taco Bell Using Agentic AI?

Building on its Byte By Yum AI platform, which already uses AI to take customer orders at drive-through windows, Yum plans to deploy virtual restaurant managers. However, it also says that it doesn’t believe they will replace human management jobs.

As reported by Reuters, a video demonstration of the concept involved a character referred to as Byte AI Restaurant Coach. The character explains that they can help a restaurant manager track crew attendance and plan shift patterns. It also makes suggestions like altering opening hours to match market conditions and even taking over at the drive-through window.

Although clearly not presented as a market-ready product, it’s a strong indicator that the business – the world’s largest franchise operator – understands agents are the next step.

Yum’s platform is currently used by several other multinational brands that sit under their umbrella, including KFC and Pizza Hut. It offers both customer-facing (e.g., drive-through ordering) and internal AI (e.g., shift management) applications.

Putting an agentic layer above this, bringing all the data, insights, and (critically) actions together in order to understand their impact across the entire business, is the goal here.

about:blankClick to accept marketing cookies and enable this content

Virtual Management

Taco Bell’s vision for Bytes illustrates the progression of AI from very trivial, routine tasks to those involving thinking, planning, and decision-making.

This mirrors a trend we’re seeing across all industries, as business leaders look for use cases beyond automating very routine tasks like processing customer orders, or customer assistance chats.

The aim isn’t to replace managers—the types of decisions that AI agents can make about managing teams and human behavior will still require human oversight.

Instead, it’s about creating virtual management assistants to step in and offer advice and guidance when it’s needed or when there are tasks that it can clearly do better than a human.

Developing AI that can interface with our lives in this way, rather than just waiting for us to ask it questions or tell it to do specific tasks, is a problem we need to crack before we approach AGI.

In fact, it’s critical to some of the most hyped AI use cases across industries, like virtual healthcare assistants, teaching assistants and legal advisors.

So Yum doesn’t actually believe its platform will lead to a reduction in the number of fast-food management jobs. Instead, it aims to augment existing human workers so they can spend more time on more valuable work.

Faster Food?

Competition to leverage technology first is always fierce in fast food, and there are missteps. McDonald’s scrapped a pilot project involving AI drive-through agents last summer because it wasn’t great at getting orders right.

This is just one of many AI use cases at the Golden Arches. One of the most recently announced involves a collaboration with Google Cloud to deliver predictive maintenance for restaurant machinery. Heavily invested in AI, it’s highly likely it will soon find other opportunities to put agents to work.

Globally, fast food is a trillion-dollar business with a reputation for rewarding players who are the first to exploit emerging technologies.

Robot chefs, AI-driven personalized menus and voice-ordering systems are all set to revolutionize the industry in the coming years. It isn’t difficult to see agentic management platforms serving as the intelligence backbone for this infrastructure.

While the industry has already weathered significant transformation, the arrival of AI agents could be the industry’s biggest shake-up yet and will certainly test which companies are set to dominate in the AI era.