Whether it’s cutting costs, driving productivity or creating more time for strategic initiatives, AI is delivering real business impact. But despite these scaling benefits and the excitement that surrounds them, we must remember that AI is still in its infancy.
It needs constant sense-checking and human supervision because, just like people, it makes mistakes, sometimes with significant consequences.
Remember the AI hiring company that was fined for rejecting applicants based on their age, or the parcel-service chatbot that insulted its own company? These are not isolated incidents, of course, and as AI adoption scales, the need for effective policies and guardrails will only increase.
The new hybrid workforce: human and machine
One topic garnering particular attention this year is that of the amalgamated human and machine workforce: employees working together with AI “colleagues” to improve efficiency and execution. In fact, this is fast becoming the new hybrid workforce that business leaders need to recognise and invest in – and while many of us may have baulked at this idea just a few years ago, it’s now a reality.
What’s more, with two-thirds (66%) of companies already reporting measurable productivity gains from agentic AI, the future of work is set to look very different, very soon.
So how can organisations optimise this new workforce demographic to benefit both people and the bottom line?
Continuous performance management
We’re not suggesting managers conduct one-to-one performance reviews with AI “employees”. But as digital team members become the norm, it’s inevitable that their development and performance will have an increasing impact on business outcomes.
This is why AI agents, much like human employees, require continuous and constructive feedback, course correction and positive reinforcement.
If they are to learn what’s useful and what’s not, they need to be told “that was a helpful answer” or “no, that wasn’t relevant”. Over time, an organisation’s ability to do this well will also create competitive advantage as traditional teams fail to contend with the human and machine model.
The criticality of AI upskilling
A big part of this calls for AI upskilling, and it comes in two parts: firstly, if employees are to reap the benefits of AI, they need to be upskilled in how to use it optimally and responsibly. The problem is, just 16% of the UK workforce has had any form of AI training over the past 12 months, which suggests innovation is already outpacing capability.
To catch up, organisations must work to close their most critical capability gaps through skills-based learning that connects the dots between upskilling and what the business is trying to achieve.
Most employees will also need to become skilled in managing and developing AI “colleagues” – a task that presents new and unfamiliar territory. In many respects, this may be easier than managing a human. An AI team member does not have feelings so it cannot be offended, for example. There are no difficult conversations that require planning and careful navigation.
Managers can be very direct with their feedback and tone. And unlike their human counterparts, AI agents have an unlimited memory capacity, which means they retain every bit of knowledge that’s ever fed to them. They forget nothing.
Human enablement: the foundation for success
But for all these differences, there are also several similarities that span the management of humans and AI agents.
We know, for example, that human employees do not need more learning content. What they really need is the right content, at the right time, mapped to the skills they need to perform in their job role. This also applies to AI “colleagues”. The goal is not to cram AI agents with all manner of knowledge and content but rather to provide them with the most relevant and credible content.
Once again, this is where human expertise is vital for knowledge-sharing, teaching and coaching agents towards high performance. In fact, it is this human enablement that forms the foundation of success – and for the organisations that get this right, the long-term value and return on investment will be immense.
A cycle of capability-building
Perhaps the best thing about a well-managed human and machine workforce is that it will eventually begin to support human upskilling. It works like this: if humans invest in developing their AI “colleagues,” those colleagues will begin to teach and support other human employees who can then re-invest their knowledge, creating a cycle of capability-building.
Now imagine this at scale. The unparalleled efficiencies and productivity gains, and the growth opportunities. This is now the future of work and it represents an unbeatable task force.
Five key takeaways to help your organisation build a high performing human and machine workforce:
1) Educate managers on the importance of developing AI colleagues.
2) Encourage company subject matter experts to commit to the cause by lending their
knowledge and continuous feedback.
3) Use clear communication and feedback to help AI agents understand their job role and
organisational context.
4) Regularly test the agent’s knowledge and performance to ensure accuracy and
mitigate bias.
5) Keep a human in the loop to supervise and provide the subjective analysis
that AI cannot yet deliver.
Nelson Sivalingam is CEO and Co-Founder of HowNow - the AI-powered learning and skills platform. He is also the author of award-winning book 'Learning at Speed', and co-host of the popular 'L&D Disrupt' podcast. Nelson has been recognised by Virgin Media Business as one of the top 30 young innovative founders in the UK, and recently featured on Bloomberg's Entrepreneurial Mindset documentary.
- Nelson Sivalingam
- Nelson Sivalingam
