If an AI can scam old people out of their retirement I dont understand how the drive through attendant isn’t just replaced with an AI yet. I know easy to trick and all that but that’s the one job most people hate at fast food. Add like 5 speakers so people can place 5 orders at once and then have the person go to work making food instead of taking orders.
I saw a story recently where a guy spent some time with a customer service chatbot, and ended up convincing it to give him 80% off, and then ordered like $6000 of stuff.
LLMs just don’t produce reliable/predictable output, it’s much easier for the user to get them to go off the rails.
Aren’t there also tons of studies and math that show/prove they cant differentiate between instructions (e.g. from the company) vs data (e.g. that guy’s messages)?
Of course in any other application, keeping instructions and data separate is very important. Like an SQL injection attack is when you’re able to sneak instructions in where data is supposed to go, and then you can just delete the entire database, if you want. But with LLMs the distinction doesn’t exist.
McDonald’s briefly had an AI run their drive thru. Apparently it got a lot of complaints, but honestly it massively improved my local McDonald’s order accuracy and speed. It was significantly better than the extremely shitty employees they normally have working the line.
Taco Bell had it, too. I have never actually completed a purchase with an AI.
Years ago, I adopted a personal policy of driving off as soon as a restaurant attempted to upsell. The Taco Bell AI always attempted to upsell me. 100% reliable on that offensive behavior. But what really and truly pissed me off was that even if I told it “No” or remained silent to its query, it always added the item to my order.
I’m happy to tank their KPIs as “reward” for their AI bullshit.
If an AI can scam old people out of their retirement I dont understand how the drive through attendant isn’t just replaced with an AI yet. I know easy to trick and all that but that’s the one job most people hate at fast food. Add like 5 speakers so people can place 5 orders at once and then have the person go to work making food instead of taking orders.
Oh they’re trying. Experiments have made so many mistakes.
I saw a story recently where a guy spent some time with a customer service chatbot, and ended up convincing it to give him 80% off, and then ordered like $6000 of stuff.
LLMs just don’t produce reliable/predictable output, it’s much easier for the user to get them to go off the rails.
Aren’t there also tons of studies and math that show/prove they cant differentiate between instructions (e.g. from the company) vs data (e.g. that guy’s messages)?
Yes, I believe that is the case.
Of course in any other application, keeping instructions and data separate is very important. Like an SQL injection attack is when you’re able to sneak instructions in where data is supposed to go, and then you can just delete the entire database, if you want. But with LLMs the distinction doesn’t exist.
McDonald’s briefly had an AI run their drive thru. Apparently it got a lot of complaints, but honestly it massively improved my local McDonald’s order accuracy and speed. It was significantly better than the extremely shitty employees they normally have working the line.
Underpaid and mistreated employees don’t make for employees that go above and beyond.
Taco Bell had it, too. I have never actually completed a purchase with an AI.
Years ago, I adopted a personal policy of driving off as soon as a restaurant attempted to upsell. The Taco Bell AI always attempted to upsell me. 100% reliable on that offensive behavior. But what really and truly pissed me off was that even if I told it “No” or remained silent to its query, it always added the item to my order.
I’m happy to tank their KPIs as “reward” for their AI bullshit.