AI in 10
The most important AI story—explained in 10 minutes.
Every day, I break down the biggest AI story in just 10 minutes - what it is, why it matters, and how you can actually use it. No tech jargon, just AI made simple.
AI in 10
Uber just burned through a year's AI budget in 4 months
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Referenced Links:
Uber slaps hard cap on employee AI tool spending after burning through budget
Enterprise AI scaling costs force budget constraints
AI workplace adoption creates unexpected expenses
AI tools strain enterprise budgets across tech industry
Uber
Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.
Welcome to AI in 10. I'm Chuck Getchell, and every day I break down the biggest AI story in just 10 minutes. What it is, why it matters, and how you can actually use it. Uber just burned through its entire annual AI budget in four months, and now they're rationing how much AI their employees can use. I'm Chuck Getchell. This is AI in 10. What happened? Why it matters, what you can do with it. Let's go. So here's what went down earlier this week. Uber rolled out AI coding assistance and productivity tools to help their engineers work faster. You know, the usual story about autocomplete your code, generate tests, write documentation, ship features faster. Classic AI productivity pitch, except nobody at Uber apparently did the math on what happens when thousands of engineers start hitting AI models all day long. Every single time someone asks the AI to complete a line of code or generate a test case, that costs money. It's not like traditional software where you pay once and click forever. These AI tools charge per token, per request, per generated line. And when you scale that across an entire engineering team, well, let's just say finance got a very unpleasant surprise. Four months in, Uber's finance team looked at the AI bills and basically said, we need to talk. The company had blown through what they budgeted for the entire year, not the quarter, the year. This is like planning to spend $50 a month on coffee and then realizing you've been hitting up the fancy espresso cart six times a day. So now Uber has implemented what they're calling token caps, basically AI rationing. Each employee gets a monthly limit on how much AI they can use. Once you hit your cap, you're back to writing code the old-fashioned way. Now, this isn't Uber giving up on AI. They're not pulling the plug or declaring it a failure, but they are admitting something a lot of companies are about to discover. AI at scale is expensive, really expensive, more expensive than most people planned for. Here's why this matters if you're not writing code at Uber. This is probably the first domino to fall, not the last one. If your company has rolled out AI tools, whether that's coding assistants, writing helpers, customer service bots, whatever, there's a decent chance your finance team is having the same conversation Uber just had. Wait, we're spending how much on AI tokens this month? And when that conversation happens, the easiest solution is usually caps or limits. Maybe you get 10 AI summaries per day instead of unlimited. Maybe the writing assistant gets turned off for certain teams. Maybe the really powerful models get reserved for high priority projects only. The point is, if you've gotten used to having AI tools at work, don't assume they'll always be there in the same way. Companies are learning that unlimited AI usage can turn into unlimited bills pretty quickly. But here's the thing about this whole situation: it's actually revealing something important about how AI is being adopted in most companies. A lot of leadership teams said everyone should use AI without really understanding what that costs or how to measure if it's worth it. So employees started using AI for everything. Not just the high-value tasks, but the random stuff too. Generate a quick email, rewrite this paragraph, help me debug this function. Make this meeting summary a little cleaner. Each request by itself is cheap, maybe a few cents, but multiply that by thousands of employees doing it dozens of times a day, and suddenly you're looking at serious money. Uber's solution of caps isn't necessarily wrong, but it does create an interesting dynamic. Now, employees have to ration their AI usage. They have to think about which tasks are actually worth spending their monthly tokens on. Which honestly might not be a bad thing. Instead of using AI reflexively for everything, people might start using it more strategically. Save the AI for the tasks where it really adds value and handle the simple stuff yourself. This reminds me of when companies first started giving employees smartphones and said use them for work without thinking about data plans. Suddenly everyone was streaming videos and downloading apps on the company dime until someone in accounting noticed the bills and implemented data caps. Same pattern, different technology. So, what does this mean for your actual day-to-day work? A few things to pay attention to. First, if your company is experimenting with AI tools, start paying attention to how much you're using them and for what. Because if caps come down at your workplace, you want to make sure you're using your allocation for the stuff that actually makes you more productive, not just the stuff that's convenient. Second, this is probably a good time to get comfortable with both AI tools and non-AI alternatives. Don't let yourself become completely dependent on AI for tasks you could handle manually if you had to, because there might come a day when your AI budget runs out halfway through the month and you still need to get work done. Third, start thinking about return on investment when you use AI. Not in some formal corporate way, but just is this AI request actually saving me significant time or helping me do something I couldn't do otherwise? Or am I just using it because it's there? The employees who are going to thrive in this new environment are the ones who can use AI strategically rather than habitually. Now here's something practical you can do today. Whether you're using AI at work or just for personal stuff, start tracking your usage for a week. Not obsessively, just keep a rough count. How many times are you asking Chat GPT to help with emails? How often are you using AI to rewrite text? How much are you leaning on coding assistance? Just get a baseline sense of your patterns. Then ask yourself, which of these uses actually save me meaningful time or produced better results? And which ones were just nice to have? Because if companies start implementing caps like Uber did, you want to know which AI tasks are worth your tokens and which ones you can handle just fine without help. For example, maybe using AI to help structure a complex presentation is worth it, but using it to rewrite simple emails probably isn't. Maybe AI is incredible for generating test data, but overkill for basic code comments. The goal isn't to use less AI, it's to use AI more intentionally. And honestly, that might lead to better outcomes anyway. When you have unlimited access to something, it's easy to get lazy about how you use it. But when you have to choose, you tend to focus on the areas where it really makes a difference. This whole Uber situation is also a preview of what's probably coming in the consumer world. Right now, most AI tools either have generous free tiers or flat monthly pricing. But if the costs stay high, we might start seeing more usage-based pricing for consumers too. Your photo editing app might give you 10 AI enhancements per month. Your writing tool might cap how many documents it'll help you with. Your coding assistant might have limits on how many suggestions it provides. That's not necessarily bad, but it means the AI, everything all the time world that some people are imagining might be further away than we thought, at least until the costs come down significantly. The bigger lesson here is that we're moving from the try AI for everything phase into the use AI strategically phase. Companies are starting to realize that just because you can automate something with AI doesn't mean you should, especially if the costs add up quickly. This is actually healthy for the industry. It forces companies and individuals to think more carefully about where AI adds real value versus where it's just a nice-to-have convenience. And for people in the workforce, it creates an opportunity. The employees who learn to use AI efficiently, getting maximum value from minimum usage, are going to be more valuable than the ones who just use it constantly without much thought. Uber's AI budget crisis is really a story about the transition from experimentation to optimization. Every new technology goes through this phase. First, everyone tries it everywhere. Then reality sets in, costs get scrutinized, and people start focusing on the use cases that actually matter. We're watching that transition happen with AI right now, and if you can get ahead of it, learning to use AI strategically rather than reflexively, you'll be in a much better position when the caps and limits start rolling out at your workplace. That's today's AI Intent. If you want to go deeper and learn AI with a community of people just like you, join us at aihammock.com. I'll see you tomorrow, my friends.
