The cost of implementing this system vs paying a teenager minimum wage makes it not worth, and probably not worth it for a while. A lot of use cases for generative AI seem to miss that you can get humans to do them for pretty low costs and installing a new AI system is quite expensive and potentially error prone.

>The cost of implementing this system vs paying a teenager minimum wage makes it not worth

That's a math conclusion so lets show our work

Minimum wage is $16/hr in California + employee overheard so lets say $25/hr. With 3 people covering shifts at the drive thru, that's $219k per year.

Now for this system, lets look at off the shelf retail API prices to get a conservative guess. $0.006/min or $8.64 per 24 hours to listen & transcribe the customer, then add TTS + GPT-4o for $50 per million tokens in and out (thats way more than they need for a day). So that's $21k per year.

So back of the napkin math says it's 10x cheaper to implement this system. Unless you're referring to the corporate R&D budget. But that's a one time cost.

I'm not sure I believe that math though. The person taking the order is also the person responsible for the drinks and the payment. Most of the inside is already done through app and kiosks. Drive through also has the ability to pick up app orders. The drive through order taking isn't the big cost here. You would still need that person to hand food, prep drinks. The real time is in prepping the order not taking the order. Reason for the line at the drive through is the meal prep not the ordering process. If AI slows down the ordering process in any way it is actually a net negative.

Many McDonald's have 2 windows in the drive-true, one for payment and one for food & drinks.

Exactly. A textbook example of using first principles for reasoning rather than just saying 'The cost of implementing this system vs paying a teenager minimum wage makes it not worth' - the equivalent of cavemen reasoning.

> Minimum wage is $16/hr in California + employee overheard so lets say $25/hr. With 3 people covering shifts at the drive thru, that's $219k per year.

The part that you can get an AI to do, you can just as easily outsource to Indonesia or somewhere.

Where are teenagers getting paid minimum wage for these jobs though? Where I live fast food jobs now start at $15/hr, not because of any laws, but because you simply can't really get anyone cheaper than that.

There have been tons of articles recently about how rising fast food prices are causing people to consider it a "luxury", and a huge part of those rising prices are rising wage costs. Where I live all the McDonald's now have kiosks inside for orders and fewer cashiers. I absolutely think shaving labor costs is something all fast food joints are interested in.

I do agree with the other comment that drive-through workers multitask (scooping fries, filling drinks, etc.), so I'm not sure how much savings you'll get just from order automation.

The labor costs are not what's causing the price rises, the CEO have said so themselves on earnings calls. Also the costs of implementing the system at all restaurants when you have stuff like Kiosks already makes it overkill. They've already automated as much of the experience as is practical with AI's current limitations.

>The labor costs are not what's causing the price rises, the CEO have said so themselves on earnings calls.

Labor costs at every segment in the supply chain are a factor. Warehouse workers, truck drivers, produce producers, ect. The price of everything that McDonald's (and everyone else) uses is higher in part because of these greatly increased labor costs across the board.

They are but if you look at McDonald's gross profits charts you'll see that profit has grown quarter over quarter since 2021 [0]. Prices would have increased less if it was purely caused by labour cost increases across the supply chain.

It's the usual greed.

[0] https://www.macrotrends.net/stocks/charts/MCD/mcdonalds/gros...

Perhaps just a turn of phrase, as they are very close to minimum wage and some of the lowest paid jobs you can get.

The point is that you can hire real live humans for not that much money.

It becomes super valuable when the error rate is lower than humans though. I think this will be the main driver for applications like this.

Yeah, but from what I can tell, there are some fundamental issues to be solved before we get an error rate anything like that with Generative AI.

I agree but would like to share an anecdote. There are some orders that I've found nearly impossible for humans to get right until I prompt engineered them.

"Two double quarter pounders with cheese" will either be two quarter pounders with cheese, or one double quarter pounder with cheese. With humans I have to order "a double quarter pounder with cheese", wait for them to enter it, then "another double quarter pounder with cheese".

It only makes sense to pay extra for AI if you scream “WE’RE USING AI” at the top of your lungs to investors, hoping they’ll throw more money at you.

Perhaps the worst thing that could happen is if the AI slows down customer throughput. If it saves money on labor but makes orders take longer that could be a big problem.

Also all the bad publicity / videos on social media hits their corporate image of competence so it's also damaging their brand.