Not at all. This is no different than those times an employee cannot hear or understand a person and asks them to pull up to the window. Is that also an "awful customer experience"? Because I can tell you right now, humans are not right 100% of the time taking drive-thru orders.

Further, you're not even considering the responsiveness. AI begins taking the order immediately, where employees often have a several second delay before they can get to you, and that's assuming they don't flat-out tell you to wait and then finally take your order a minute later. It's been my experience that the AI ordering (at Bojangles which I talk about in another comment) is faster and easier than all but the very best employees (so in other words on average it's better).

I'm not sure if this is worse, or just different, than people being prompted into suicide by assholes online, or in person, etc. There was a point in time where it seemed to be a fairly regular occurrence that someone committed suicide because someone egged them on, or was just extremely insensitive to the reality of the person's situation, online.

This seems like a tempest in a teapot. All this system does in any anomalous condition is dump the customer over to the drive-through employee to take the order manually like before.

It shouldn't be difficult to define escape conditions for this. If quantity is over X (for ANY item), or if it's been more than 30 seconds and no item has been added to the order, etc.

A system like this doesn't have to work 100% of the time, just 95% of the time to take some load off employees to (supposedly) allow orders to be filled faster.

Bojangles (primarily a southern chain in VA, NC, SC and GA) have been using AI at the drive throughs and it works great. I don't know how immune it is to trolling, because I'm not an ass, but it works great for normal people.

It should be a criminal case, they wilfully sabotaged equipment that someone has bought and paid for.

Yes exactly this, post it anonymously via some site hosted in a far away country well outside the jurisdiction of DMCA or similar rules. It's not like they will be able to do anything about a site hosted in russia or china.

... and always have been completely bedazzled on why MS Word even has a business case. How this piece of software could gain the market let alone survive to this very day is a mystery to me.

Of course it stops being that one I encounter regular users, but objectively there is no real reason for MS Word to even exist beyond some fringe niche scenarios.

I have this contract I'd like you to sign....

That was what killed real D&D. (Well, really it's successors did.)

I don't know the licensing policies of the "stock image" sites, but it's not unreasonable that they might have a license to use stock images, rather than that particular stock image. In which case it could have been left in by mistake. I believe lots of companies use lots of stock images for a variety of different purposes.

Now if they licensed that particular image, then he's almost certainly lying rather than just probably lying.

I'm not saying you could do that with one of the public versions of ChatGPT. But I think an LLM could be custom trained to do it. I also think it would take a *LOT* of time and effort. Probably enough to make this a silly idea.

Previous comments have been drawing analogies to Black Mirror, but this "idea" goes back much further...

...This is an episode of Max Headroom (US version).

Specifically, S02E02: "Deities." A company claims to be able to bring past loved ones back to "life" as an AI, for a modest recurring fee. But Bryce (the creator of Max Headroom) opines they can't possibly have the compute power to do it, as it requires a large mainframe just to run Max's highly flawed, glitching bust.

Wouldn't surprise me if the "visionaries" behind this saw that episode, and saw an opportunity to fleece gullible rubes.

And how is that different from the Washington Post or the New York Times, or most of the mainstream newspapers?

It *is* possible that his silly explanation is the truth. That they just accidentally left a "fill in later" image in place. (There's a reason text users frequently use "lorem ipsum" for that kind of purpose.)

ChatGPT making up things and incorrectly following policy would merely be in line with the current editorial pool, who ignores policy when it suits them and fabricates material if an article has any kind of political content.