Social media is full of memes about AI blundering tasks that we find easy. But does that mean AI is worse than us at everything? Even the worst critic of AI will probably agree that AI does some things more efficiently than humans. This criticism stems from our view point of seeing AI as a competitor rather than what it actually needs to become: an assistant. Imagine you are hiring an assistant to help you with daily tasks. Would you ask, “Can they do the job as well as I would?” or “Are they making me more efficient?” I’d argue the latter is a good measure of the assistant’s contribution. The same goes for AI.
The goal for AI isn't to be better than us but to help us save time and money while trying to deliver human-quality results
. A big reason why we see AI, especially GenAI, as a threat is to defend a very fundamental human need: financial and job security. Yes, AI will shuffle the job market. We have seen this shift before. Remember when the internet democratized knowledge? Jobs then shifted to Skills — the ability to apply that knowledge. Now, GenAI has both knowledge and the capability to apply it. What does that mean for our job prospects? Will we need to get more creative, beyond just skills and knowledge? Can everyone make that transition? The future of jobs depends on what we build with AI. Right now, venture capitalists favor AI tools that automate human work because they’re safe bets. We need AI tools that create new industries or make existing ones easily accessible — think autonomous environmental monitoring or healthcare and investment assistants that motivate more people to participate. Let’s talk about GenAI travel planners for a minute. We need to be skeptical of them. Our tests found some hilarious (and sometimes alarming) suggestions from GenAI: taking a 4-hour hike right after a 5-hour hike, leaving a baby at a restaurant to catch a Broadway show (seriously, people, stop feeding internet your dark thoughts, AI is learning), or crossing a river without a bridge. Do you remember the early days of GPS? Getting stuck on a dirt road in a foreign city with no clue where you were? It was not fun. But we learned to trust GPS over time. GenAI travel planners are at that same early stage right now. Most even come with a warning not to follow their recommendations blindly. They’re fantastic for quickly pulling together information for your trip. Research shows US travelers spend about 9 hours, spread over 45 days, researching their trips. So, the real question is: “Can a GenAI travel planner gather trip information faster than I can?” As these models improve, errors will decrease. Until then, embrace their quirkiness and use them for research, but with a bit of caution.