AI Agents: our new colleagues

I write Future Work/Life to help you build great teams and careers. If you find it interesting, please share it!

And if you'd prefer to receive this newsletter in your email inbox first every Thursday, join 5000+ subscribers on Substack (check out the details in my profile).

AI agents are the new buzzword in tech.

Every major tech company and hundreds of AI-cash-laden startups are racing to release their ‘thing’. And with each new ‘thing’ comes the big talk about how agents will transform how we work. Because agents aren’t just another productivity tool. They represent the start of something much more profound.

We're about to start working alongside non-human colleagues.

And OpenAI 's Sam Altman predicts that in 2025, we'll see:

“Systems that people look at - even people who are sceptical of current progress - and say 'Wow'."

What’s the big deal with agents?

Ok, let's start with defining what an AI agent actually is.

HubSpot co-founder Dharmesh Shah puts it simply:

"Software that uses AI and tools to accomplish a goal requiring multiple steps."

Today's AI tools need constant human direction - they're like helpful assistants who need instructions for every task. Whereas AI agents can handle complex work independently, like experienced team members you can trust to get things done.

That’s why Jensen Huang , CEO of 英伟达 - whose chips are powering the AI revolution - envisions his company evolving into :

"A 50,000-employee company with 100 million AI assistants."

Let me show you what this could look like in practice across three different areas.

B2B Marketing

B2B teams have led the adoption of new automation technology over the past 10 years, so they seem like a good place to start.

Let’s start with their goal. Because if you take a step back, it's not about running campaigns or hitting targets - although that’s where they’re focused day-to-day. They exist to connect the right people at the right time with the right message. Meaning that in the end, that person becomes a customer.

How about with AI agents in the mix?

Instead of individual AI tools handling separate tasks, imagine an entire AI team working in concert. One agent monitors market signals and competitor moves around the clock, instantly feeding insights to another that refines targeting strategies. A third crafts and adapts content in real-time, while others analyse response patterns and refine the process.

This isn't science fiction.

Companies like thirdweb are already using a coordinated system of agents to manage their sales and marketing outreach and nurturing. And even a large complex organisation like McKinsey is showing what’s possible by:

“Creating an agent that will speed up the client onboarding process. The pilot showed lead time could be reduced by 90% and administrative work reduced by 30%.”

Learning & Development

Learning and development is another area that’s seen huge advances in digital transformation over the past decade, with AI already improving accessibility and personalisation.

The limitation L&D professionals have is time. Because step back from tracking course completion rates and the need for new, ever-relevant content, what are they really trying to achieve? As Arist's Michael Ioffe put it on the Future Work/Life podcast, it's about enabling people to grow and adapt throughout their careers.

“The only metric that matters is [learner] outcomes and performance.”

What’s the difference in an AI agent world?

Instead of one system trying to serve everyone the same way, you have a coordinated team of AI specialists. One agent analyses individual progress and organisational needs to craft personalised journeys. Another automatically updates materials when processes change. Others track real development against business outcomes and connect learners with the right expertise at the right moment.

This is already starting to happen. Arist will release their agent, called ‘Teammate’, in Q1, while 微软 already has now released the capability to create autonomous agents in Copilot Studio. Apparently:

“Organizations like Clifford Chance, McKinsey & Company, Pets at Home and Thomson Reuters are already creating autonomous agents to increase revenue, reduce costs and scale impact.”

Legal

The law, with its complex processes and high stakes, feels ripe for transformation.

Law firms traditionally operate like something akin to a very expensive library - repositories of knowledge (some in physical libraries, some in people’s heads) accessed when needed. Their teams spend countless hours:

• Monitoring regulatory changes across jurisdictions
• Reviewing documents for potential issues
• Managing case workflows and deadlines
• Coordinating between multiple parties

But what do legal teams provide beyond document processing and billable hours? They help organisations navigate risk and ensure compliance in an increasingly complex world.

So, what could the AI agent future of law look like?

Instead of relying on individual lawyers to monitor everything, you have a legal team powered by AI. The moment a case file arrives, multiple AI specialists spring into action. One agent parses requirements and deadlines. Research agents instantly search through legal databases. Others analyse case patterns, draft arguments, and ensure compliance. All working together. All the time.

Harvey , a well-funded AI start-up, is working on this using OpenAI’s models, explaining:

“The next generation of models, alongside agentic systems and Uls, will allow Harvey to build collaborative agents that work with legal experts on their hardest problems.”

What about the humans?

So, the obvious question here is, “What about the humans?” After all, I said we’d be working alongside these things, not be replaced by them.

Well, as that quote from Harvey suggests, the opportunity here is reimagining what humans could achieve with the right support.

• How can the best marketers join the dots more effectively?
• How can L&D professionals support people developing the right skills faster?
• How can legal professionals use their analytical and communication skills to reduce risk and create more opportunities for clients?

And rather than just asking, "What can AI do to save money and increase productivity?" ask, “What could we do differently, with more time, supported by a tireless collection of super-smart assistants?"

That’s enough for today.

Next week, I’ll explore how leaders should approach this transformation, why we should value an entrepreneurial mindset, and some practical steps to start rethinking work in your company.

Enjoy the rest of your week.

Ollie