AI Thinking

What are humans good for?

What should humans ... and what should AI do?

Katie Lips, CEO and Founder of Rebel Futures/22 September 2026
What are humans good for?

I like to be contrarian… instead of worrying humans are so basic they’ll be overtaken by a Large Language Model, I think the opposite. Humans are absolutely amazing at all sorts of things; more things that AI can’t do, than things AI can do. In the world of work, however, the day to day for many humans has been taken up by repetitive, low level tasks, that AI is just going to be better at doing. If we stop seeing AI as a threat, though, and see it as an opportunity, we can free ourselves from the ‘basic’ stuff and get on with what we really should be doing: the knowledge work, the judgment, the actual intelligence.

Asking yourself “what are humans good for?”, is vital if you want to use AI well and elevate yourself and your teams from the basic stuff AI is good at. Failing to ask that question, by the way, means your AI initiative is more likely to fail.

I went to an event last week at Syrena, a startup building what they describe as an “AI employee for relationship work”. Their product, Ren, finds the people you need to meet, works the network, chases the follow-up, and gets you introduced. What it doesn’t do is have the meeting; that part is for you.

Founder Aziz Ben Allala even opened his presentation with a question I’ve been considering (professionally) for a while: the “What are humans good for?” question. Actually, Aziz asked “What should AI do?” and “What should humans do?”

And not to be provocative, but to highlight a fundamental point we can’t ignore when designing AI systems: that there are some things humans do best, there are some things AI does best, we need to stop competing, and start designing for the split.

I think a lack of an answer to this fundamental question is why AI pilots have such a high failure rate. Why? Because in a standard enterprise AI rollout, an AI assistant tool gets deployed, people get access, a training session happens, and adoption is measured (often with terrible adoption rates). A lack of training, or a lack of effective change management are usually blamed… but for me there’s always a larger problem: no one’s figured out how human roles will change. No one wants to accept or acknowledge that humans will now need to do different things; that they can now do different things.

AI has simply been added to a human system, and people have been left to work out for themselves what that means for their role. Most of them conclude, quite reasonably, that it means their role is under threat; of course they would. So it follows that they use the new AI tools as little as possible. If someone had asked the “what should AI do?” and “what should humans do?” questions, they’d likely have a better solution that allowed each to flourish.

Clear Human / AI delineation works well in some sectors already. Doctors used to spend a significant portion of their day writing up consultation notes, and updating patient records. It’s necessary work, but it doesn’t use the Doctor’s smarts. Nowadays, AI scribe tools like Tortus and Accurx leverage AI to transcribe consultations, and update records in real time. What should the AI do? Exactly this: take the strain, write the notes, update the records, crunch the numbers. While the doctor does exactly what the doctor should do - exercise their expert judgement, offer the human touch, connect with their patients. Doctors don’t get upset they’re no longer having to write notes.

Call centres are also innovating with AI by replacing the human-not-necessary work; and keeping all important human contact where it matters. Before AI, call centre agents would perform a range of tasks that AI can better perform today. They’d answer the phone, identify a problem and then find information to solve the problem. Nowadays, AI can do all that easy stuff for them enabling them to focus on the tricky conversations and the customers who really need a friendly ear. In this way, the role of a human call centre agent is elevated. These guys become more important and more valuable; not less so.

Whether we’re talking professional networking, customer service or clinical care, I think the argument’s the same. Which is… never ask “how can we use AI to make what we do today more efficient?” Instead, start by asking “what are humans (and AI) good for?”, and build your tools around that split.

Rebel Futures helps business and technology leaders navigate AI-driven disruption and win in the AI era.

Katie Lips is CEO and Founder of Rebel Futures. She helps organisations grasp the opportunities at technology inflection points, from the early web and pre-iPhone mobile to cloud-based SaaS and now AI. She has held senior roles at EY and Aimia, and was Head of Digital Policy at Which?. Katie holds a Harvard Kennedy School qualification in Leading in Artificial Intelligence and is a Founder Institute graduate.

Related: explore our work on AI Adoption.

base44
Edit with Base44