As AI becomes more widespread in the workplace and society, what impact will it have on your job, your life and the world around you? If AI can take on more and more of the tasks people perform at work – and do them more efficiently – where does that leave human beings?
In Taking the Anxiety out of AI, Sameer Rawjee explores the nature of both human and artificial intelligence, asking what AI can and can’t do, what tasks will always remain human, and what a future of human-AI collaboration might look like. Whether you’re already using AI or just starting to explore its capabilities, this book is your essential guide to navigating the future with clarity and purpose.
READ MORE: The New Age of Nutrition: Why AI Will Change How You Eat
Sameer brings over a decade of experience helping people find meaning in times of great technological and societal change. As the founder of Google’s Life Design Lab, he’s advised thousands at companies like KPMG, Investec, SoundCloud and IDEO, as well as top institutions such as INSEAD and London Business School. His insights into how we work – and how we thrive – are more relevant than ever as we enter a new era of computing.
When Machine Brains Become Like Human Brains
Sam Altman, CEO of OpenAI/ChatGPT, wrote a blog post five years after the organisation started, to reflect on his journey. To summarise the success of his work in three words he said, ‘Deep learning worked.’ In other words, his team figured out how to help computers learn just like humans, and that meant they were on track to create a perfect artificial brain.
So what did he mean exactly? Well, computers have always just taken instructions from humans. A human will program an algorithm or formula into a computer to tell it what to do when users click buttons. An engineer tells a computer to give you the answer to a maths problem when you use the calculator on your iPhone. The engineer inputs basic maths rules such as 1 + 1 = 2 and then the computer learns how to add any size of numbers together. It also does this much faster than humans ever could. This means that computers can ‘think faster’ than us – and that is already amazing. But what if computers could also ‘learn faster than us’?
What if computers could teach themselves new languages (as happened at Meta in 2017 when Mark Zuckerberg decided to shut down their new self-learning AI), or ask new questions that humans have never asked, or see patterns in data and make recommendations, or read all the world’s information and create new insights, pieces of art or strategies. Wouldn’t that be phenomenal? Ultimately a large part of our creativity just comes from reorganising information that is already out there in different ways. So if a computer has ‘all the information out there’, then it could also be creative. Sometimes creativity comes from your intuition or ‘the gods’ but other times it’s just about seeing patterns in the world and being the first to package and sell them. That is why ChatGPT is so successful: it has become this computer that can take all the world’s knowledge and find new patterns in there to give you the creative answers you want. Moreover, this can be achieved in seconds. Ideas that might have taken you weeks to piece together can now be sourced, organised and packaged faster than ever. And that is an incredible feat.
Hence, the new era of computing is the era in which machines evolve from just thinking faster to learning faster – and that’s what is changing everything. That’s what deep learning is really about, it’s fundamentally what AI is really about, and why Tesla, Oracle and Microsoft are spending in excess of $100 billion to make computers that can learn incredibly well and may perhaps later even act on our behalf. This means that the evolution of computing would be thinking, learning and acting. Once we get to ‘acting’, we are effectively saying that we trust how well a computer thinks and learns, so we may as well let the computer act on what it knows without our intervention. This is where customer service chatbots have evolved to already, and where Sam Altman and his team believe there is an equally viable path to create the next CEO chatbots that will run billion-dollar companies with no help from humans. That is effectively Silicon Valley’s interpretation of AGI, or when computers will be considered to be as smart as humans.
How Human Brains Transcend Machine Brains
An incomplete understanding of intelligence will always lead to an incorrect interpretation of AGI. Most people understand intelligence as only the cortex and don’t know much about the other intelligences and their role in helping the species evolve. Devoted logicians (people who swear by logic no matter what) in particular, who think everything can be attributed to their brains alone, will soon see that there is more to humans than just copying the brain.
Sir Ken Robinson said in his TED Talk titled ‘Do schools kill creativity?’ that ‘intelligence is not distinct’. All your intelligences are working together in unique ways: emotions, intuitions, intellectual processes and the five human senses all come together to inform new ideas and create new patterns among information – thereby producing new knowledge. Machines can only work with what they have and make that better. They cannot find new ‘inspirations’ from nature, meditation, emotional connection, therapeutic discovery, taste or smell. All these experiences are often fundamental in channelling new ideas to the brain so that new logic, science and frameworks can be created.
READ MORE: Your Best Year Yet! How AI Can Double Your Productivity
Moreover, machines may be able to reorganise philosophical ideas in ways that surprise humans; so, if you ask ChatGPT an existential question, you may still find an incredibly useful answer. But AI cannot spiritually reflect on how ideas best fit together in any given circumstance. Sometimes real human intelligence involves moral reasoning, empathy and ethical decision making in complex, ambiguous situations – often where there are no clear answers. In high-stakes situations, these types of decisions are made case by case, through a thread of discussions by groups of people attempting to arrive at outcomes that both ‘look’ right and ‘feel’ right. This happens every day in tough business decisions, policymaking, law, corporate governance, and even manufacturing or distribution. Sometimes the choices or decisions we make are so illogical that it’s hard to think that a computer, which is purely logical, could also help us arrive at those same great conclusions.
Still, the future is not about human versus machine. If it typically took 12 months for 50 people in government to make a policy decision, AI may reduce that time to 12 weeks. AI speeds up the basic phase during which we would typically get tired, mentally bogged down, hungry, sick, busy with kids, or caught up in our traumas and all the other dilemmas that slow down our mental processing power. Now all we need is one good morning to run a great set of questions through AI and we’ll acquire all the basics we need. Now we are not spending months trying to find, organise and package information from which to make a decision. With the right questions and the right AI assistant, everything you need is available to you within minutes, and you can now spend the rest of your time discussing, reflecting and deciding. That is what humans are for.
Excerpted from TAKING THE ANXIETY OUT OF AI by Sameer Rawjee. Copyright © 2025 by Sameer Rawjee. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
