AI IS GOIING TO UBER EVERY ONE OF US
And it is not about driving cars
8/12/20264 min read
Have you ever thought about driving for Uber? Regardless of whether you like the service or agree with its business model, you should know that one of the functions of AI is to make the relationship between the hiring party and the service provider closely resemble the Uber model. This isn't strictly about the nature of the employment relationship, but primarily about compensation.
Let’s look at what this means... To maximize profitability, Uber created an algorithm designed to charge the customer the highest possible price while paying the driver the lowest possible amount. To achieve this, it decoupled the two: paying more for a ride doesn't necessarily mean the driver gets paid more. Quite the opposite—the ideal scenario for this model is to collect the maximum amount and pay out the minimum.
Obviously, doing this without AI is highly complex; most companies would end up spending more time managing the system than is feasible. However, with AI, this is not only possible but desirable—at least from the companies' perspective. Let’s be clear: this practice isn't unique to Uber. It has long existed in the airline industry, where you pay more when seats are scarce, especially for last-minute bookings. Essentially, it operates like a reverse auction: you are presented with a price and decide whether to pay it, with the cost tending to rise as the time of service approaches.
Uber simply added two extra layers: it decoupled the price paid by the customer from the driver's pay, and it introduced a dynamic coefficient that changes during the ride—one of the most controversial features in recent times. Notably, Uber didn't invent this concept either: in a traditional taxi ride, you pay more and more as traffic worsens. Therefore, variable pricing isn't inherently absurd, yet the implementation was poorly received and eventually hidden from the customer's view.
By enabling these time-sensitive micro-calculations—as seen with Uber—AI creates the potential for a service-hiring model that easily straddles the line of ethical acceptability. First, because it goes beyond analyzing immediate conditions—such as driver availability, trip location and characteristics, and the willingness of both parties to accept the ride—to consider broader contextual factors: your propensity or need to pay more or accept less, based on financial data and behavioral history.
Let me elaborate... Imagine you are applying for a job. Your profile looks good: you have a clean credit record, a solid professional history, and so on. However, you have a massive credit card balance (the company purchased a customer database from major card issuers), you just bought a new car, your children changed schools, and you recently returned from a trip. In other words, that job opportunity might not be particularly great, but it could be a lifesaver—allowing the company to offer a lower salary than originally planned, set a higher target for variable pay, or perhaps propose a shorter-term contract. In short, the negotiation shifts based on your needs rather than necessarily on your competence, value, or qualifications.
And it can go even further... You might walk into a supermarket and see prices change specifically for you—much like what already happens online. But these prices could shift based on less public data, so to speak. Perhaps you are taking Wegovy and are less inclined to buy sweets, so the price drops. Or, the price might rise because you are diabetic and currently experiencing a hypoglycemic episode. How could the supermarket know this? The same way social media platforms do: through your smartphone and the data you unknowingly share, or through your banking records (showing monthly purchases of the medication pen or insulin).
The truth is, the possibilities are endless. Your hourly value could fluctuate from hour to hour, depending on the level of demand for your services. From a profitability standpoint, this is nothing out of the ordinary. From a personal standpoint, however, it is overwhelming. Ultimately, AI will make it possible to pinpoint exactly who is productive within an organization and who isn't, likely putting an end to bloated staffing structures. Today, in the vast majority of companies, there are employees whose only job is to justify their own existence within the organization. They function as elbows, controls, barriers, and detours. They are exactly like traffic tools—such as traffic lights, detour signs, speed bumps, drainage channels, and the like. They do not improve the quality of traffic, the driving experience, or the lives of those traveling; they merely create barriers and attempt to control vehicle flow to reduce bottlenecks.
In an AI-optimized operation, there is no reason to maintain this type of function. It will be better performed by AI, and this will affect people in many different ways. Some will lose their jobs, others will get promoted, but the majority will find themselves with a more limited scope. Believing that everyone will become hyper-productive with AI is like thinking that the latest iPhone will revolutionize your photos or your productivity. Tools like AI work best in the hands of those who know how to use them. The difference, in the case of intelligence, is that it can determine for itself who knows how to use it and who doesn't.
When you hear that AI won't wipe out jobs and that everyone will have more free time, consider this: either it’s a lie, or everyone in the future will be an Uber driver or a life coach. Probably both.