I think in this case - where an editorialized title is better representative of the article, but the published title is too long - it’s usually left to someone like you to give a suggestion, publicly, which is then seen by @dang or their kith, who can then actually make spot adjustments. I personally approve of your editorialization, but that’s worth naught but a canker in a hedge
Lets just break the lynx users for once, and see who complains. Change it to 81. Small enough to be invisible to most users, but when it breaks, someone will come up and shout. Better than going to 100, 132, 140 or 280.
It seems like the article's unspoken consensus is if they hired people (ideally local) and paid them a living wage it would go back to working well enough like it did before they attempted the bot.
I wonder why no one brought that up as a solution?
I am wary of stories like this, since they seem to appeal to what people want to be true rather than what is necessarily happening.
There are a lot of businesses quietly and conservatively implementing AI and seeing real results - it's just the crazy Jerry Springer blowups you hear about!
Yes, but that only some businesses do that flys in the face of the breathless, borderline psychotic boosterism a lot of us tend to hear quite a lot of, that AI is gojng to revolutionize everything.
Directly customer facing is a tough one, since a lot of the time chatbots are used to just fob off customers that eat up a lot of time (old people complaining etc).
I think the best experience I've had is when companies offer an MCP server that lets me get up and going with their product very quickly. Sequenzy is one such example I tried, very AI-first and works well with Codex.
If you're talking about stuff that isn't customer facing, there's all kinds of tasks that can and have been automated.
As a simple example, I run a retro games business and we've vibecoded up various controller/cartridge testing apps that have dropped the time it takes to test a box of product by around 50%. I'm planning to spin up a few Hermes agents soon to watch the prices of things and suggest improvements to our listings (e.g. noticing we haven't soldas many copies of Super Mario 64 as we should, and checking our keywords/pricing).
Obviously puts in big work with SaaS and stuff too.
EDIT: actually, Shopify's built-in agent works really well and is directly customer facing. It's a smart model (I think either a Kimi or Qwen as a base) that seems to have been fine tuned on their product itself.
You can ask it stuff like "what products have seen a decrease in sales" or "what is the value of my next payout" and it just answers them.
Before you had to manually write DB queries for this stuff, or more realistically just ignore it.
MCP is just API, and we have been there already and we know what happens: Businesses realise they can’t extract maximal value if they don’t own the interface so they recant the API.
AI driven individual incentives for decision makers. Comp goes up for showing off AI, not hiring people.
Lots of “thought leaders” who push the solution but will be gone for the consequences. Next folks (usually those who remain) clean up the mess to no fanfare.
>Comp goes up for showing off AI, not hiring people.
That's funny because a few years ago it was the reverse. people were saying how big tech was bloated, caused by managers that felt their clout was proportional to how many people were under them.
You get what you measure. Both can be wildly irrational depending on incentives and target outcomes.
Would an AI solution have been in place in this context if testing revealed this outcome we’ve arrived at and a data driven decision was made to hire instead? Or was no testing done and we’re testing in prod? Sure looks like testing in prod or simply ignoring metrics specific to a quality customer experience. Like investigating a crash site, we have to ask what series of events and decisions have led us to this incident.
A pharmacy tech actually counts the pills if they move them from Bottle A to Bottle B, it is not particularly time consuming, and it isn't a particularly skilled position. Starts at $14 an hour around me and tops out around $23.
Machines cost a ton of money and need to be maintained?
As those two variables change, I'm sure they will replace the dispensing function for pharmacists. In fact this already does happen at a lot of pharmacies these days. Machines put the pills in bottles, then a pharmacist hands the bottle to you.
They are in large setups. Including mail order and bulk dispensing places (like Walmart pharmacy—most pills are dispensed at central locations and delivered to the store prepackaged).
I could imagine a machine like this would cost something on the order of $100k, require ongoing maintenance and wouldn't really solve any problems/bottlenecks at your typical local pharmacy.
It's not like you could replace the pharmacist with a machine.
Impressive: they managed to use LLMs to recreate the CVS pharmacy experience, where you get random phone calls from random pharmacies offering to refill a prescription you don't need refilled, or better yet, they just call you to tell you it's already refilled.
A paradoxical situation here is that they are probably not using the frontier models because compliance and other stamps of approval but this packaged approved AI is making silly mistakes in reading forms
What I've noticed is that AI excels at tasks that are self-contained within a specific logical framework, like coding, writing, and mathematics. But conversation isn't self-contained and requires too much contextual information, which is why AI chatbots and counselors don't seem to perform well.
Ironically, coding and mathematics are about simplifying the complexity of the world and reducing it to logical procedures. That's exactly where AI performs well. But in any profession that involves friction with the real world, current AI still seems to fall short.
Let me predict how Kinney management is going to react to this fiasco: They're going to call it a PR problem and say they haven't done a good enough job of selling the benefits to the customers. So they'll double down and tell the customers they're holding it wrong.
That will of course just piss off the customers more.
The right solution which will restore customer trust is to rip out all the AI shit, call it a failed experiment, and go back to the old system. They'll never do that because then the VP who thought up this garbage won't get his bonus.
This assumes a ton of HIPAA data doesn't get sent up to some cloud and then breached. If that happens, people will start going to prison. HIPAA enforcement does not fuck around. Or at least it didn't before AI and Elon took over everything.
You should post the entire title.
reply