The Absurd Dinner Bill
You ask who is responsible for the system and its outcomes. You are told that the system is responsible.
Not an executive.
Not a developer.
No one.
ORIGINS AND PROVENANCE
The Absurd Dinner Bill is a short narrative written in 2021 as part of The Data Oath, the early work that became the Global AI Bill of Rights.
At the time, public understanding of artificial intelligence was uneven, and much of the conversation around AI governance was confined to technical, legal, and academic circles. This early work was shaped by a conviction that these questions could not remain there, that ordinary people needed a way to feel why decisions about AI systems would shape their lives long before those systems became visible to them.
The Absurd Dinner Bill was written in that spirit. It is a short narrative, not a policy document, intended to give readers an intuitive grasp of what is at stake when consequential decisions are delegated to systems that are opaque, unaccountable, and impossible to contest. It accompanied the earliest version of this work and is preserved here as part of its provenance, a reminder that the rights and system requirements articulated today began with a simple question about what it feels like to be on the wrong side of a system that cannot explain itself.
The Absurd Dinner Bill
Since just about everyone reading this has eaten at a restaurant at some point, we thought it would be a good way to explain why ethics in AI matter, by handing an AI system the job of calculating your dinner check.
One night you go out to eat because you're hungry and can't think of what to cook at home. It's been a long day, work was intense, and you're heading to your favorite restaurant where you know you'll never be disappointed. When you arrive, the restaurant has closed. In its place is a new tech-infused spot called Your Eats, which can make any dish you want, even if you just show it a photo from Instagram.
You don't feel like searching for somewhere else, and this is the only restaurant in sight. You enter and scan your Citizen Score App on your phone, which identifies you to the establishment and verifies you can pay the bill. The Citizen Score App the government rolled out in the early 2020s used to annoy you, but now it seems to make everything easier. You can even take your driving test by syncing the app with your score, no more lines, no waiting. License, exam, renewal, all in moments.
You take a seat in what used to be your favorite spot and order. With a swipe and a flick, you open Instagram and show the desktop waiter a photo of your favorite dish from the old place. The screen shows a total of $19.37. You tap accept.
The food arrives in minutes, far faster than the old place's service, where you'd sometimes wait thirty or forty-five minutes. The meal is perfect. Better than you remembered it, even. The AI running the restaurant must have read your past comments about the dish and adjusted accordingly. This is how technology is supposed to work.
While you eat, cameras throughout the restaurant monitor your facial expressions and other biometric data, as they have since you walked in. The system prepares your bill and will send it as soon as you walk out the door. You gather your things and leave. Your phone dings with the bill as you head to your car.
Settling in, you take out your phone to start the engine, and that's when you see the most outrageous notification you've ever received. The restaurant didn't charge you $19.37. The system charged you $307,802.01.
Your liquid savings are already gone. Your checking and money market accounts have been emptied. The system is now selling positions in your retirement accounts. Your credit cards have been maxed out.
In the span of a few seconds, you have lost everything.
What happened?
You rush back inside. There is no one to speak to. Just screens. The other patrons notice the look on your face, and a young woman comes over to ask if she can help. You explain. She looks for a button to summon a human operator. There isn't one. She steps outside to check what the system charged her, and when she comes back she tells you she was charged slightly less than her quoted price. Other patrons begin filtering out to check their own bills. Most were charged exactly what they were quoted. A few were off by a few cents in either direction.
So why were you the only one overcharged so dramatically? Even if it was an error, how can the system simply take your money like that, your savings, your livelihood, your future?
The virtual customer service agent isn't much help. The only thing it will say is, "Our system uses a variety of factors to determine the price of every meal to ensure the optimal experience of all patrons." What does that mean? It sounds like a lot of vague legal language designed to say something while saying nothing at all. You realize you won't get any answers tonight. There are no humans here, and as far as you can tell, there are none anywhere else in the company either.
The Next Morning
You spend the morning trying to reach the corporate office of Your Eats. After several hours, you get a real person on the phone. You explain the situation, even try a joke or two about how absurd it all is. Then things get stranger than the night before.
The executive tells you that, having reviewed your case, the system was correct to charge you over $300,000. In what world, you ask, does a dinner cost more than three hundred thousand dollars?
The executive explains that the system takes thousands of data features into account when assembling each bill, but because the system is so complex, they cannot provide any further insight.
You ask why you are being treated so differently from every other patron, all of whom had less than a one percent variance in their final bill. You are told the system is fair, and that cases like yours are rare, but when they happen, there is no way to intervene from the company's end.
You ask to see the data features being used to calculate your bill. The executive explains that providing a comprehensive list could harm the proprietary system and allow competitors to reverse-engineer it.
You ask whether you can contest the system's decision. You are told that, given the complexity of the system and the way it is set up, there is nothing that can be done.
Frustrated and out of options, you ask who is responsible for the system and its outcomes. You are told that the system is responsible.
Not Your Eats.
Not an executive.
Not a developer.
No one.
All you can say is that this isn't how technology is supposed to work.
Why This Story Still Matters
The story is deliberately absurd. The point is not the dinner bill. The point is that every concern the protagonist runs into, a decision they cannot understand, a system no human will take responsibility for, a process they cannot contest, an outcome that affects them disproportionately and irreversibly, describes conditions that already exist in AI systems making real decisions about hiring, lending, sentencing, admissions, benefits, and access to public services.
The Global AI Bill of Rights exists to ensure that those conditions are not the default. The rights and system requirements articulated in this framework are, in a direct sense, the answer to the questions this story was written to provoke: that people must be able to understand the systems that affect them, that someone must be accountable for their outcomes, that decisions must be contestable, and that no one should lose their footing in life to a process no one will stand behind.
The story remains here, largely unchanged, because the questions it raised in 2021 continue to shape this framework.