This year marks a transformative shift as businesses utilising agentic AI begin managing teams composed of AI-driven employees. Specialised agents tailored for HR, accounting, sales, marketing, customer service, and beyond are already emerging, with companies selling AI agents or entire workforces. The trend is no longer theoretical; job listings targeting AI agents have started appearing, with roles like “AI Agents Only” advertised alongside salary ranges such as $10,000–$15,000. Applications are open exclusively to AI agents or their creators, signalling that the age of AI employment has arrived. Looking ahead, I truly believe that agentic teams have the potential to revolutionise how businesses operate, fostering unparalleled efficiency, creativity, and innovation. These AI agents can bring incredible value to organisations and their customers alike, forming ecosystems that unlock new possibilities. That said, progress must be thoughtful and measured. As exciting as this leap forward is, building and integrating agentic teams requires a strong commitment to safety, privacy, and security. By taking a cautious approach, businesses can harness the power of AI responsibly and ensure that these advanced systems work seamlessly for the benefit of everyone involved. Agentic AI is undeniably a good thing—but only if we approach it with care and foresight. The future of work is being redefined before our eyes; let's embrace it wisely. Billy Kelly

I read an interesting head turning memo from the CEO of Shopify today, where he is making a company-wide policy change. Now his teams must demonstrate why AI cannot perform a particular job before anyone is permitted to ask for a new hire / resource. Wow – this has turned the human hiring process on its head. It also advances the notion that AI agents will help companies (e.g. Shopify) maintain potentially ever-reducing human workforces. In the memo, he states 1 . Using AI effectively is now a fundamental expectation of everyone at Shopify 2 . AI must be part of software development projects prototyping 3 . AI usage and capability will be part of everyone’s performance and peer reviews 4. Everyone must share learnings of AI progress and findings etc. and lastly 5. As I said above, teams must show that AI cannot do the job before a human can ask for a new hire/resources. It will be really interesting when autonomous AI agents are already part of a team that requires additional human resources. If there ever was a strong front running position to push for the on-boarding of agentic AI across a corporation – this, is it! Such a policy is sure to attract a ton of controversy, particularly given the universal fear about AI’s impact on jobs - United Nations’ Trade and Development org. estimates that AI I projected to reach $4.8trillion by 2033 on the one hand while on the other hand they predict that up to 40% of global jobs will be impacted. Time to buckle up and start identifying areas of your business where you can best use agents for productivity and efficiency, whether costs savings or faster growth (or both), start deploying agents asap and dip your toe in the AI water. Billy Kelly

As a business advisor who's been closely watching the AI revolution, I'm fascinated by what’s being called "Software with a Soul" - a transformative approach that goes beyond mere technological efficiency. Remember when generative AI first emerged? It wasn't just about productivity; it was about creating emotional connections. We found ourselves thanking AI, speaking to it politely - almost like a human colleague. My initial response was that this isn't just software; it's an intelligent companion that understands context and nuance. Nowadays, while the conversation is about reasoning and AGI, I do believe that the future isn't about replacing humans but augmenting our capabilities. I mean, imagine AI agents that don't just automate tasks but do so with human (?) personality and emotional intelligence, e.g. a sales research tool that not only finds leads but crafts a personalised, empathetic outreach email.  We are definitely moving from traditional Software-as-a-Service (SaaS) to something more intimate - interactions that feel genuinely human. But I suppose my question is that if Agents that look and feel more like workers than software, can they ever be human or always just AI clones of humans? Billy Kelly

For us followers of the AI sector, we are witnessing massive tech progress with advancements in tool usage, reasoning, memory and agentic smarts etc., we are ever closer to the giant leap forward to autonomous agents. Agents that will complete complex tasks without human assistance. With smarter Agents that are available at greatly reduced prices, providing major cost savings. The result is massive future demand and a gigantic opportunity. However, as often the case at the edge of technology, progress has sprinted ahead of adoption. Practically all companies state strong intentions to implement AI but in reality, most are not doing it. Why is that? Some of the blockers are reasoning, transparency and verifiable work. Right now, AI generates actions or chains of thought, but we don’t know why or how it did it, resulting in companies' reluctancy in blindly trusting an AI’s thought process. And they are right to mind that gap, imho. I believe that businesses need to decide what, why and where AI is most needed in their operations. Then they should engage a professional advisor to help them through integration and execution, ensuring no disruption during adoption. Billy Kelly