Artificial Intelligence (AI) has the potential to revolutionize industries, streamline operations, and uncover insights that were once impossible to access. Yet, many organizations find their AI initiatives underwhelming, falling short of the promised value. Why does this happen, and how can you avoid these pitfalls? Let's dive into some of the most common reasons AI projects fail and how to turn things around. 1. Misaligned Use Cases One of the most common issues is treating AI like any other IT system. AI thrives in scenarios where it can solve complex, dynamic problems, not in rigid environments best suited for traditional systems. If your AI use case doesn’t leverage the technology’s unique strengths, like pattern recognition, natural language processing, or decision-making under uncertainty, it’s likely doomed from the start. We've seen companies try to use AI to try and automate spreadsheet updates. There is already a solution for that: Macros! How to fix it: Begin by understanding what AI is (and isn’t) capable of. Define use cases that align with your business objectives but also take advantage of AI’s strengths. AI is not a plug-and-play solution; it’s a tool that requires thought and strategic application. Remember even though you're carrying a hammer, not everything is a nail! 2. Overly Ambitious Projects AI evolves rapidly, and what’s cutting-edge today may be outdated in 6 months. Attempting long term AI projects is a recipe for disaster. Some things you'll struggle to do today may be simple in a few months time. How to fix it: Focus on small, incremental projects that can be completed in less than four weeks . Quick wins will not only showcase the value of AI but also provide valuable learning opportunities to refine and scale your efforts over time. 3. Too Many Cooks AI is new and sexy and everyone wants to be involved, we get it. But as soon as you start to increase the number of stakeholders, management hierarchies and areas involved things tend to grind to a halt. How to fix it : Successful implementations require collaboration between a small group of empowered people. Think of it like prototyping: let them iterate small solutions and then share what works to see what can add value at scale. 4. Insufficient Data Quality and Quantity AI models are only as good as the data they’re trained on. Poor-quality, incomplete, or biased data can lead to inaccurate or irrelevant results. Many organizations underestimate the effort required to clean and prepare data for AI projects. Ask yourself the question: have previous projects have failed due to poor quality data? How to fix it : Invest time and resources in data preparation. Audit your data for quality, completeness, and relevance. If gaps exist, prioritize data collection initiatives before diving into AI development. Remember you may have good quality data which is just poorly structured in which case AI may well be part of the solution! 6. Unrealistic Expectations AI hype has led many organizations to expect miracles, often without a clear understanding of the technology. Unrealistic expectations can result in disappointment, loss of stakeholder confidence, and premature abandonment of AI initiatives. How to fix it: Set clear, achievable goals for your AI projects. Communicate openly about the capabilities and limitations of AI to manage stakeholder expectations effectively. And again, look for those small quick wins, they will build confidence and excitement. Conclusion: Think Big, Start Small, Iterate Fast AI implementation isn’t about building the flashiest solution, it’s about solving real problems effectively. By narrowing your focus, embracing rapid iteration, and empowering teams, you can set your organization up for AI success. Remember, the key to unlocking AI’s potential lies not just in the technology itself, but in the strategic and thoughtful way it’s applied.  At pm ² we specialise in the rapid development and deployment of bespoke AI Agents so that you can quickly realise the benefit to your business. Talk to us if you want to learn more.

It's rare that a week goes by without a tech company releasing an update to their AI tool and boasting about how much better it is than the previous version. The technology is improving so quickly, it can be hard to keep up. So what does that mean for your business? Are you already using AI to improve your productivity, just embarking on your AI journey or unsure where to start given the pace of change? Wherever you sit it's important to have a view on how AI is going to transform the working environment and how you you will need to adapt. And if you're not ready to go all-in on AI just yet, one way to dip your toe into the water is to use On-Demand AI Agents . These will be bespoke agents developed for you and could be perfect if you are waiting to fill vacancies or are struggling to find the right candidates in certain roles. It could be that you simply need some manual invoice reconciliations done and you're waiting to recruit a new accounts clerk. Or it could be you need someone to trawl through your paper based contract and pull out all the key SLAs information. You can think of using On-Demand AI Agents as similar to hiring a temp, but you're getting it at a much cheaper price! When thinking of what Agents you need, it's helpful to think about how they will fit into your business: 1. Integration into Routine Tasks AI agents can handle repetitive, labour-intensive tasks like data entry, invoice matching, document translation and summarising. These tasks, traditionally done by employees, will be more efficiently managed by AI, freeing up human workers for more strategic and creative endeavours. 2. Enhancing Productivity and Flexibility AI agents can be available when needed, allowing companies to scale their workforce quickly and cost-effectively. Businesses will benefit from increased productivity and flexibility, as they can deploy AI agents to handle peak workloads or specific projects without long-term commitments . 3. Collaboration with Human Employees AI agents can work alongside human employees, providing support and augmenting their capabilities. Employees will focus on tasks that require human creativity, emotional intelligence, and complex decision-making, while AI agents handle the more routine aspects. 4. Customization and Specialization AI agents can be customised to meet the specific needs of different businesses. Industries like finance, healthcare, legal services, and customer support can particularly benefit. For example: Finance: agents can handle tasks like reconciling transactions, analysing financial data, and generating reports Healthcare: agents can assist with medical record management, appointment scheduling, and patient data analysis Legal Services: agents can review and summarize legal documents, conduct legal research, and manage case files Customer Support: agents can handle initial customer inquiries, provide information, and escalate complex issues to human agents. The bottom line is that using On-Demand AI Agents enables you to get up close and personal with the technology and its capability without any long term commitment: perfect in an environment which is developing so rapidly . At pm ² we specialise in the rapid development and deployment of bespoke AI Agents so that you can quickly realise the benefit to your business. Talk to us if you want to learn more.

You've probably been seeing the term AI Agent or Agentic AI a lot more recently. Many people believe that 2025 is going to be the year of the AI Agent . But what are Agents, how will they develop over the course of the year and how do you maximise their potential to improve business performance? Read below to find out more: AI Agents and Agentic AI are like super-smart assistants that can help us with various tasks and make our lives easier. They've been around for a while but we're now hearing about them more and more. Their time has come due to the progress being made in AI Technology, the decreasing cost of related infrastructure and the increased availability of user AI tools from the big technology companies such as Microsoft, Google, OpenAI and Meta. The simplest way to understand an AI Agent is to think of a simple task like checking an invoice. An AI Agent can be built to do that task, so in that example it can check an invoice, identify errors and categorise them. Another AI Agent can be built specifically to deal with each category of error. For example, one category may require an email to be sent to a supplier, another category may require an invoice to be resubmitted etc. A process can be broken down into many tasks and individual agents can be assigned tasks to complete the process. Once an AI Agent has been built it can be used again and again and in multiple different processes. You may already have interacted with an agent and just don't know it yet (many chatbots use AI Agents to identify and process different types of requests). Even if you haven't it's likely that you may interact with one or many in the foreseeable future as they will become common place in many areas of business: Personal Assistants : Imagine having a personal assistant who knows your preferences and can help you with daily tasks like scheduling, shopping, and even reminding you to take your medication. These AI agents will be more intuitive and personalised. Healthcare : AI agents will assist doctors by analysing medical data, helping diagnose diseases earlier and suggesting personalised treatment plans. This could mean better healthcare for everyone. Education : AI agents will help teachers and students by providing individual learning experiences, tutoring, and even grading assignments. Making education more tailored, effective and accessible. Customer Service : Businesses will use AI agents to provide faster and more efficient customer service, answering questions and resolving issues around the clock. An AI Agent never sleeps and never needs to take a break. Automation : AI agents will automate repetitive tasks in various industries, increasing productivity and allowing humans to focus on more creative and complex work. Imagine at some point in the future you have a Virtual Assistant that knows your daily routine. It wakes you up at the right time, reminds you of your appointments, orders groceries when you're running low, and even suggests recipes based on what you have at home. This virtual assistant can also help you with work tasks, like drafting emails or finding information quickly. This Assistant would be built on a network of AI Agents, each of which are optimised to do a specific task. 

How AI Consulting will impact Small to Medium-Sized Businesses AI Consulting is a service that helps businesses use artificial intelligence (AI) to improve their operations. More companies, particularly those in the SME sector are turning to AI to make better decisions, improve customer experiences, eliminate inefficiencies and stay ahead of the competition. AI consultants are important because they help these businesses understand and use AI technologies. This article will explain how AI Consulting can help small to medium-sized businesses succeed. Understanding AI Consulting AI Consulting involves working with teams within the business to find the most appropriate Use Cases for AI. To this end, AI consultants do several key things: they assess a business's needs and create AI strategies and they help implement AI solutions that can automate and optimize processes. AI consultants understand data and analytics, which means they can analyze large amounts of data to provide useful insights; finding opportunities and solutions that the business will not have seen before. They have technical skills in machine learning algorithms, data processing and LLMs (Large Language Models). They stay up to date on AI trends so that they can offer the latest and most appropriate AI solutions. AI Consulting is a team effort, with consultants working alongside different teams in a business. Benefits of AI Consulting for Small to Medium-Sized Businesses AI Consulting brings several benefits to small and medium-sized businesses (SMEs). It allows SMEs to access AI technologies without needing in-house experts or resorting to costly software and consultants from the likes of Google, Microsoft and similar. This is cost-effective because it avoids heavy investment in new infrastructure. AI Consulting helps improve efficiency by automating tasks and streamlining processes. It can also greatly enhance customer experiences through personalized interactions. AI-driven insights support better decision-making, as consultants provide data-backed recommendations. By adopting new technologies quickly, AI Consulting helps SMEs stay competitive. AI solutions can also scale with a business, supporting growth over time. Challenges and Considerations in AI Consulting But like any type of consulting it's best to go in with your eyes open. It's crucial to align AI solutions with your business goals, don't get side lined by great technology which doesn't have a clear business benefit. Data privacy and security are also important concerns that consultants must address, depending on the models and data sources being used. Moreover, integrating AI with existing systems requires careful planning and clear communication between consultants and the business. Over-relying on AI can be risky, so human oversight remains important in AI processes. And remember, AI is a fast moving technology, there are new developments literally every day. So, it's important that whatever solutions you choose have a clear upgrade path when increased performance is available. Future Trends in AI Consulting It's likely that emerging trends in AI will shape business futures and therefore AI consulting is going to be key to help businesses navigate through the ever evolving AI landscape. If you don't do it then your competitors will. Can you afford to fall behind!

As AI evolves we are really starting to see how its true capabilities will change the software landscape forever. It is nothing short of a revolution and the thought leaders in this are the heads of the big tech companies: Google, Meta, Microsoft, Amazon, OpenAI, Tesla/X. A fascinating interview with Microsoft CET Satya Nadella last week described his thoughts on the far reaching changes AI will bring. His prediction runs along these lines: Most systems used by companies interact with core business data: customer, sales, operations, employees etc. These interactions typically fall into 4 categories known as CRUD (Create. Read. Update. Delete) AI Agents can be created (in fact they pretty much create themselves given clear instructions in English) with the capability to conduct CRUD on most data no matter where it is held or what format it is in Therefore .... whole SAAS systems can be replaced by a force of connected agents developed specifically for your business. There will no longer be any need for the one size fits all approach where compromises on functionality need to be made And the thing is, AI Agents are already here and getting better everyday. What's even more interesting, Satya thinks that even tools like Excel will become unnecessary; remember this is the CEO of Microsoft saying this . It's a fascinating short interview and it's worth taking a look, here is a nice version with commentary . It's clear that this is going to be a head scratcher for many companies with large IT Infrastructures and existing investments in multiple internal and 3rd party systems. Many of these are likely to get bogged down in the as-is and lose the opportunity to get to the new to-be state. Let's face it, if Microsoft are talking about it, they're planning for it and so are their peers. Fundamentally this is why AI strategy and decision making needs to be led by the CEO. These are business decisions and the correct execution can enable a company leapfrog over competitors maintain or achieve market dominance. Worrying about where you are now and what systems you have will not help.

Elon Musk's active involvement in the recent U.S. presidential election has not only influenced the political landscape but also holds significant implications for the future of artificial intelligence (AI) policy and development. He's got himself into the cockpit and it's likely he'll have the Pilot's seat when it comes to all things AI and technology. Musk's Political Engagement and AI Advocacy In the lead-up to the election, Musk emerged as a prominent supporter of Donald Trump, contributing substantial financial resources and leveraging his social media platform, X (formerly Twitter), to bolster Trump's campaign. His America PAC (Political Action Committee) played a pivotal role in mobilizing voters, particularly in swing states, underscoring Musk's commitment to shaping political outcomes. Beyond his political activities, Musk has been a vocal advocate for responsible AI development. In August 2024, he endorsed California's Senate Bill 1047, which mandates safety and accountability measures for large AI systems. Musk acknowledged the bill's potential to upset some stakeholders but emphasized its importance for AI safety. Potential Influence on AI Policy With Trump's election victory, there is speculation about Musk's potential role in the new administration. A petition by Americans for Responsible Innovation has called for Musk to be appointed as a special adviser on AI, highlighting his expertise and advocacy for responsible AI development. If Musk assumes an advisory position, it could lead to a more cautious approach to AI regulation, balancing innovation with safety. His support for AI safety legislation suggests he may advocate for policies that ensure AI technologies are developed responsibly, potentially influencing federal AI regulations and funding priorities. However, some have commented that his positional support for regulation has more to do with his falling out with Open AI front man Sam Altman (Musk was an original co-founder of OpenAI) and a strategy to delay it's progress so that his own AI product, called Grok, can catch up. Impact on AI Industry and Innovation Musk's involvement in AI extends beyond policy advocacy. His company, xAI, launched the Grok AI chatbot, positioning it as a competitor to existing AI models like OpenAI's ChatGPT, Microsoft's CoPilot and Anthropic's Claude. Grok aims to offer real-time responses and a unique user experience, reflecting Musk's commitment to advancing AI technology. Grok also benefits from being able to train on all the content that goes through X (formerly Twitter) on a daily basis. However, Musk's dual role as a tech entrepreneur and political influencer raises questions about potential conflicts of interest. His companies, including Tesla and SpaceX, could benefit from favorable policies, leading to concerns about the impartiality of AI regulations under his influence. Conclusion Elon Musk's active participation in the recent U.S. election was a gamble that could pay off handsomely. He now has a potential role in shaping AI and technology policy in the largest market in the world. His advocacy for responsible AI development, combined with his entrepreneurial ventures, positions him as a key figure in the future of AI innovation and regulation. As the new administration takes shape, Musk's influence could lead to significant developments in AI policy and the use of AI and technology to drive increased efficiency and effectiveness within government.

The recent UK Autumn Budget 2024 has introduced a rise in the National Insurance Contributions (NICs) that employers pay by 1.2 percentage points, combined with a decision to lower the threshold for when firms start paying, to £5,000 from £9,100 per year. This increase in NICs means higher payroll costs for businesses, potentially leading to reduced hiring and/or slower wage growth. Smaller businesses, in particular, might feel the pinch more acutely as they often operate with tighter margins. But fear not, by adopting AI, businesses can not only mitigate the impact of increased NICs but also position themselves for long-term growth and sustainability. AI agents don’t attract National Insurance, sleep, take time off or need to be paid and they can be more productive than humans. Specifically AI Agents can address: Cost Efficiency : With higher payroll costs, businesses are more motivated to find ways to reduce expenses. AI can automate routine tasks, such as payroll processing and invoice matching, leading to significant cost savings. Workforce Optimization : AI can help optimize workforce management by predicting staffing needs, scheduling shifts, and even identifying areas where human labour can be reduced without impacting productivity. Compliance Monitoring : AI can continuously monitor and ensure compliance with new regulations, reducing the risk of penalties and the need for additional compliance staff. Data-Driven Decision Making : AI can analyse large datasets to provide insights into business operations, helping companies make more informed decisions that can lead to cost savings and efficiency improvements. Innovation and Competitive Advantage : Investing in AI can give businesses a competitive edge by improving their operational efficiency and enabling them to offer better services or products. By staying informed and adapting to these changes, businesses can position themselves for success in a competitive market. Embracing AI technology and leveraging its benefits can help businesses thrive in a rapidly evolving business landscape. Stay ahead of the curve and embrace the future of work with AI agents.

At work we are surrounded by data, huge amounts of data in fact! Some of it is digital data stored in databases, spreadsheets, knowledgebases and computer documents. But much of it is often legacy hardcopy data like 3rd party contracts, SLA agreements, employee contracts etc. The question is how much of this data can you turn into knowledge? Can you access the multiple different strands of data that you have in a meaningful way and use the results to effectively enhance business performance. If you can't then you should consider using AI to help speed things up. WHY? Well, in the same way that AI powered internet searches such as Copilot or chatGPT bring back richer and more relevant information than a simple Google search, AI can do the same when scanning your business data. And it doesn't matter what format the data is in, AI can read it for you. This means it gives you the capability to link and interrogate different sources of data from one place that you couldn't link before. What's more, AI can act on the output of the information and automate tasks. As an example, here are two business problems with AI powered solutions. Website Self Help no longer helping : Company A has built an extensive website with lots of self-help advice for customers to self serve but there is now so much data on the website that customers try a couple of searches, don't find what they are looking for and end up calling the support number. So now the website that was supposed to reduce calls is starting to create them! Solution : Use AI to build a chatbot that responds directly to the question the customer types in with the specific information that they are looking for and continually learns customer behaviour to generate a daily top 10 list of problems/solutions that can be displayed. Invoice Matching in a mess : A company has to manually match a proportion of payments to invoices and purchase orders before it can ship the products from stock. It is taking a lot of time, delaying the dispatch of customer orders and leading to dissatisfaction. Solution : Implement the process below using AI to match, verify and then automatically trigger payments so that products can be despatched. Highlight orders were no match is found and sort these for manual intervention. Data Extraction : Use Optical Character Recognition (OCR) technology to extract data from invoices, purchase orders, and receipts. This includes key information such as supplier details, quantities, and pricing. Automated Matching : Once the data is extracted, automatically compare the information from invoices with corresponding purchase orders and delivery receipts. This ensures that the invoiced items have been both ordered and received before making any payments. Discrepancy Detection : Identify discrepancies, such as mismatches in quantities or pricing, and alert the accounts payable team for quick resolution. This reduces the risk of erroneous payments. Integration with Financial Systems : After verifying the accuracy of an invoice, automatically update the company's financial system, ensuring seamless data flow and reducing manual entry errors. Then authorise and trigger the product dispatch. Continuous Learning : Continuously learn and improve their accuracy over time, adapting to different document layouts and reducing the need for manual intervention. A company that effectively uses AI can lower it's cost base, improve customer service and identify growth opportunites, thus giving it a competitive businesses advantage. Those that don't invoke this new technology risk being left behind!

After a poor showing for Elon Musk's new Robotaxi last week his robot stole the show. The humanoid robot dubbed Optimus (presumably after the Transformer) wowed attendees at Warner Bros. studio in Hollywood. Half a dozen of them navigated their way through the crowd safely. This is no mean feat given that just two years ago the droid couldn't even make it onto stage on its own. This latest version could talk to guests, pour a drink from a tap and even play rock paper scissors. However, everything was not quite as it seemed there was a bit of Wizard of Oz involved. The droids were voiced and operated remotely by humans wearing special suits that translated their movements to the droid. When questioned, one droid said "Today I'm assisted by a human. I'm not yet fully autonomous". So not autonomous robots then, more like puppets! And with a man as famous for PR stunts as for implementation it could be a while yet before we see them in our homes like in I Robot . Although at some point in the future Must expects the robots to be available for as low as $20,000 and be selling 100s of millions a year. It's an interesting use of the word "low", $20,000 seems expensive for a product to get mass adoption. But will there be a big demand for personal robots or by the time they become reality will we be so well served by thinking appliances that can prepare and cook food, garments that don't need laundering, self driving cars etc. that we won't need them? Well only the future will show us that. There is also of course the big question which relates to the socio-political impact that robots would have on jobs and levels of employment; which is a whole topic for another day. But I don't think we'll be seeing autonomous humanoid robots in the home anytime soon.