There has been a lot of hoopla over the dangers of AI and even calls to suspend AI research for a while. If people aren’t worried about AIs taking over the world, they’re worried about AIs taking over their jobs. People are genuinely concerned, and they should be. Job loss is coming, but it’s not the AIs that are to blame. It’s the employers.

In an effort to save money employers will inevitably view AI as a tool to get something done for less. Employers already make this calculation with foreign workers and every new thing that promises to make what they do cheaper, and it make sense. Who wouldn’t want to pay less, but what will be the result of replacing workers with AIs? Will it make the products and services we consume better or worse? To understand this, we need to understand what happens when a worker is replaced with anything.

Imagine you have a worker that does embroidery. Now you buy a machine that can also embroider. By getting rid of this worker, you save money and gain speed and consistency. Seems good. You’re also able to sell more embroidery because the machine doesn’t need to sleep. Good again, but what did you lose. You lose the human, a free agent who may have just bought a shirt and thought that the fabric would be great for embroidery. You lose someone who’s capable of realizing that the embroidery no longer fits with the latest styles that people wear. You lose someone who can, through their own human experience, provide you with new ideas and insights for new embroidery designs. The machine cannot do any of this for you because by using the machine, you are committing to keeping that thing the same. So, what you really lose is an avenue for which your products and services can serendipitously improve.

This is not to say humans should be chosen over AIs either. In fact, the question of which to choose is the wrong question. The question should be, “How can you leverage AI to make your products and services better?” The question of which humans AIs can replace should not even enter the discussion because AIs cannot replace humans. There’s just so much we can do that AIs can’t. What AIs are really good for is making humans a lot better at their jobs. AIs like ChatGPT and Bingchat are great for assisting professionals with their work such as finding information, pointing out logical errors or run quick calculations, which frees up the professional to tackle more complex problems.

This is what makes AI such a powerful tool. It is a game changer with the potential to propel human intellect and creativity to new heights. It is also an equalizer, giving ordinary people access to skills and abilities that were previously available only to those who can afford it. And this is also why it is unfortunate that a lot of employers will invariably want only the non-human half of this equation. If this sounds like your employer, then it might be time to dust off that resume.

So, while the answer to whether replacing humans with AIs will lead to better products or services is complicated, history tells us that there will be both good and bad AI-powered products and services. Do you know what the difference will be between the good and the bad ones? Why does ChatGPT outperform all other AIs of comparable size and complexity? Even GPT4 doesn’t perform as well, and ChatGPT, the paid version at least, is based on GPT4. It’s because ChatGPT’s responses have been corrected. By humans.