AI is everywhere. And if you’re not using it, it can feel like you’re getting left behind. Thortok2000, another writer on the Mystro team with hundreds of rideshare hours, shares his insights on how you should and shouldn’t use AI as a gig share driver.

Let’s get the bad news out of the way right from the start: AI won’t instantly increase your rideshare or delivery earnings with omniscient precision. If you’re hoping for an easy fix that will perfectly fill your pockets and give you a shortcut to getting rich, you’re out of luck. AI will be happy to write up such a script for you, but its accuracy will leave much to be desired.

But that doesn’t mean AI can’t be helpful. While a completely perfect solution is unattainable, you can definitely get a boost from it. And here’s how…

AI is great at answering questions. It’s conversational and will give you an answer that’s plausible and makes sense, most of the time. But how good of an answer that is, will depend on what data it’s using to answer your question with.

So, where AI really shines is when you feed it lots of data first, and then ask a question about that data (making sure to use any relevant settings like Data Analysis for ChatGPT or Python for Gemini).

For example, here are some questions you could ask to get some insights on where your best earnings have come from: Where did I earn the most in the last month/year? What times of day did I earn the most? What regions gave the most tips?

You can download your data from whichever platform you drive for: Uber, Lyft, Doordash, InstaCart and more. (If you’re struggling to figure out how to download your personal driver data, don’t hesitate to reach out to the support team for that app.) Note that this data typically includes only completed and paid trips. It will not give you any data about offers you did not take, or how busy things were or weren’t during times you weren’t online or places you didn’t go.

Another fun way to use AI, especially if you can ask it hands-free while driving, is basic questions about the area or specific locations. I’ve asked AI to help recommend a restaurant to my passengers if they like a type of food I wouldn’t pick for myself, or answer questions about how expensive one is even if I haven’t been there before. If I am picking up a passenger at a business, I have asked AI if there’s any interesting trivia about that business as a possible thing to say in conversation. Asking for recent news might also spark something interesting.

Sometimes the AI can answer random questions that come up in conversation. Do you know what you call a group of flamingos? (An actual question that came up during a ride once.)

There are also AI translation apps out there if you want to better communicate with passengers that don’t share your language. While it’s not the best tool for the job, I can start Gemini from my steering wheel and ask “please ask my passenger ‘is this the correct address’ in Spanish.”

But what about boosting your earnings? Well, if you’re just starting and don’t know up from down just yet, AI can help give you a good starting point. Asking it when and where to drive will give you some decent starting suggestions based on generic advice that’s commonly shared on forums and blogs.

AI can also do a quick check for any local events in your area that may or may not bring a spike in demand, either in the recent past, present, or near future. (Make sure the AI is using live search when you ask, of course, as offline data won’t help at all.)

For gig drivers, the biggest limitation on AI use is that it has no access to the gig platforms to get real, live, concrete data that would be necessary to ensure the best decisions. AI can guess surges (time and location) based on when people on forums have talked about past surges. It can guess that a surge might happen based on when and where events happen and when they end. But it doesn’t know.

Even if the AI can maybe guess at potential demand, it will never know supply. It has no idea how many other drivers are on the road heading for that same shiny potential surge. Someone constantly asking AI “where should I go now” will find themselves pointed at hallucinated surges that never actually materialize, driving back and forth across town, wasting their gas.

The best advice when it comes to surges is to try to predict them and already be there before they happen. Getting a feel for your market (learning over time where and when the surges happen, and the kinds of things that trigger them, like which events do and don’t), will serve you a lot better than AI’s guesses.

Because really, all AI does is guess. They base their suggestions on blogs and forums, which can be informative, but will never be better than your own direct experience. Current AI language models (like ChatGPT, Gemini, and even Grok to an extent) are very heavily weighted towards never saying “I don’t know.” They will always give you their best guess, no matter how wild or made-up that guess is. They’re designed to intentionally sound confident as well, so don’t mistake confidence for competence.

If you have absolutely no idea, then go ahead and ask an AI. Some context is better than nothing. But learning your market will always be superior to anything AI comes up with. That’s going to apply to both the “where to position” and “which offers are good or bad” as well.

“But wait!” I hear some of you cry. “If the problem is the AI just doesn’t have the live data, then what about the AI that’s actually in their app? It sees everything, right? It could help me cherry-pick, improve my positioning, and earn more!”

Keep this in mind: The AI ‘tools’ in Uber, Lyft, and other apps are primarily there to help the company make as much money as possible. Not you. This is a lot like “HR is there to protect the company from you, not you from the company.”

Cherry-picking means skipping some (or even most) offers and only taking the best and most profitable. The gig platforms, however, want you to accept every single offer you get. On an individual per-driver basis, your profit model is the exact opposite of theirs. So they are not going to allow their AI tool to help you cherry-pick.

As far as positioning, the AI tool they provide is not an advanced, complex surge predictor: it is just going to repeat the surge data already on the map, nothing more. This leads you into “chase the surge” behavior they want from you when “predict the surge” behavior is smarter and earns more.

So what do those AI tools do? Honestly, not much. They are large language model (LLM) chatbots integrated with the information you could already get in the app in other ways: Where the current surges are (you could just look at the map), what events are happening or upcoming (already listed elsewhere in the app), and so on. They can answer basic questions about the service and provide basic support without the need for a human support agent. That’s all they do.

And like any other tool, it’s how you use it that matters. You need to play to the strengths of what the tool is designed to do (like using Mystro for your hands-free multi-app handling). 

Asking AI can give you a starting point, but so can asking other drivers, especially local ones. In the end, you’ll want to let your own trial-and-error experience guide you. 

Do you have any ways that you use AI as a gig driver? If so, join our Discord and share how you use AI to optimize your driving. Till then, drive safe, and we’ll see you next week in your inbox with a fresh topic.