AI needs People

There's been a lot of talk about AI taking over everyone's jobs. 

Duolingo, Microsoft, Oracle and IBM all made redundancies in 2024, based on AI being able to replace workers.

The assumptions were:

⚠️ AI can do the work of 10 people

⚠️ AI can cut labour costs by 60%

⚠️ Profits will be boosted overnight

In 2024 alone, over 150,000+ workers were laid off, with 45% of cuts directly tied to AI automation.

Duolingo cut 10% of its translation contractors when it replaced them with AI-generated content. 

But, within months, users complained about poor translation quality and cultural inaccuracies.

IBM replaced around 200 HR staff with chatbots and its AI "AskHR" system. Instead of cost savings, the company faced internal backlash and employee satisfaction dropped, forcing leadership to reinvest in hybrid roles combining AI and humans.

Microsoft laid off 15,000+ during record profits, refocusing on AI productivity. But it created new bottlenecks, product releases slowed and team morale dropped as institutional knowledge vanished.

According to a 2025 Orgvue study, 55% of leaders who laid off staff because of AI now say it was the wrong move. 

Many underestimated the need for humans. AI cannot replicate human creativity, critical thinking, and intuition.

So what's happening now? The smartest companies are seeing that the best outcome is to augment work with AI, but with humans in the loop. They are pairing human judgement with machine speed.

So when you start working with AI, apply the lessons learned here, that you can't fully replace people, no matter how good your AI solution.