The Era of "AI" Employees is Here

[Free]

The thread title is stolen from a billboard ad for Artisan, a company apparently of people who root for SkyNet to win.

AI company trolls San Francisco with billboards saying “stop hiring humans”

Company boasts "AI workers" that never complain about work-life balance.

arstechnica.com

Since the dawn of the generative AI era a few years ago, the march of technology—toward what tech companies hope will replace human intellectual labor—has continuously sparked angst about the future role humans will play in the job market. Will we all be replaced by machines?

A Y-Combinator-backed company called Artisan, which sells customer service and sales workflow software, recently launched a provocative billboard campaign in San Francisco playing on that angst, reports Gizmodo. It features the slogan "Stop Hiring Humans." The company markets its software products as "AI Employees" or "Artisans."

The company's billboards feature messages that might inspire nightmares among workers, like "Artisans won't complain about work-life balance" and "The era of AI employees is here." And they're on display to the same human workforce the ads suggest replacing.

 

[Anya Ristow]

I know of nobody who has been replaced by AI. I have heard of nobody who has been replaced by AI. When I work with AI it feels like I am dealing with a high school student who has not read the text and is bullshitting their way through an essay.

As a career computer programmer of four decades, I have automated tasks for a living. I am familiar with the concept of reducing human time spent through automation. My own time is valued according to whether I can reduce the time others spend on tasks by more than the time I spend automating them. I GET automation.

The trend of calling things AI that are not AI precedes the current AI fad by a long time. The term is finally being applied appropriately. But the tech is not as useful as is being hyped.

As far as I can tell.

Placing chat bots between customers and employees is automation of a sort, but it is negative automation. You are trading a small amount of employee time for a large amount of customer time. It may improve the bottom line, if the customer is not sufficiently pissed off by it, but it is not a net benefit in time spent, any more than placing a frustrating phone menu between customers and employees.

What AI IS doing is finding novel patterns in data that people haven't otherwise discovered. This isn't really any different than the past decades of computer science, though. As you apply computer solutions, you magnify the efforts of humans. What's changing is that it requires less and less expertise to recognize that something new has been discovered in the data.

I have a friend who came up with a problem he wanted to solve with computer programming, so he tinkered with it for YEARS. He was approaching it as a task of patching together solutions that he found online, and he kept talking of it for YEARS. I didn't say anything, but it occurred to me he'd make faster progress if he actually learned to program. And then chat-gpt came along and he thought AI was going to write it for him. But he got sidetracked. Now he's convinced he can get AI to write novels. I offered to help with his programming and he didn't take me up on it, and I offered to look at what his novel writing effort was coming up with and he didn't take me up on it.

But here's the thing about using an oligarch's AI to create a novel (no pun intended) solution. If you succeed, you've just trained someone else's AI to do it for someone with more money than you have. If you come up with something useful, they will just take it. You will know this "stop hiring humans" company has created nothing useful when nobody steals their business model.

I had the same thoughts when my ex boss started working with Amazon. If he was successful (he was not) he'd just show Amazon another working business model it can steal.

I'd laugh at all this, but I know there is A SHITLOAD of energy being burned in this AI craze.

 

[Veritable Quandry]

Mrs. Quandry was until recently a trainer for an in home elder care company.

They have just replaced the majority of HR functions, including hiring, training, and scheduling, with an AI system. Caregivers in Ohio are now hired and supervised by a computer in California.

So now you have heard of a significant number of humans being replaced by AI.

 

[Noodles]

I don't know if its AI, but I have been seeing reports of people getting banned on Instagram for basically nothing, which suggests some AI moderators making a lot of false positives.

 

[Anya Ristow]

They have just replaced the majority of HR functions, including hiring, training, and scheduling, with an AI system. Caregivers in Ohio are now hired and supervised by a computer in California.

Are you willing to name the California company? I'd love to read about it.

Lots of things that are called AI are not AI. Remember when Amazon created AI kiosks that required no human interaction? Turned out it required 1000 video reviewers in India. 70% of transactions involved human interaction. In India. That wasn't some scammy startup that is now bankrupt because they pulled a fast one. That was Amazon, and there was no consequence.

Algorithms written by humans are being called AI because there is venture capital in things that are called AI. It is the 2020's tech bubble. I'm calling it now.

Actual AI doesn't understand the concept of the number four. It spits out grammatically correct nonsense. It is a very advanced pattern recognition system, and even its creators don't understand how it works.

It CAN be used for scanning resumes (that's a pattern recognition task), but human algorithms can do that, too, and probably better.

Training? Interviewing? Resolving employee problems? Setting policy? Meeting regulatory requirements? Talking to people? If any of the "AI HR companies" are claiming to have actually automated this using AI, I want to read about it. I'm finding marketing BS that often sounds like it's written by AI. I'm betting that what is being sold as AI is actually just outsourcing, with chat bots between customers and clients.

Are you willing to name the California company?

 

[Anya Ristow]

I don't know if its AI, but I have been seeing reports of people getting banned on Instagram for basically nothing, which suggests some AI moderators making a lot of false positives.

That IS AI. Characterizing human-written text is a thing AI CAN do. It doesn't understand the text (or images), but it is able to characterize it as argumentative, pornographic, abusive, etc. Automated moderation is actually a good application. What's supposed to happen is posts that are flagged as problematic are put in a separate queue and a human (say, youtube channel owner) is supposed to review it. But humans tend not to review it, so if your post is characterized as abusive or whatever, a computer has gotten the final say, because there is no review and you have no recourse.

It makes mistakes, as you've noted.

 

[Kamilah Hauptmann]

 

[Veritable Quandry]

Are you willing to name the California company? I'd love to read about it.

Lots of things that are called AI are not AI. Remember when Amazon created AI kiosks that required no human interaction? Turned out it required 1000 video reviewers in India. 70% of transactions involved human interaction. In India. That wasn't some scammy startup that is now bankrupt because they pulled a fast one. That was Amazon, and there was no consequence.

Algorithms written by humans are being called AI because there is venture capital in things that are called AI. It is the 2020's tech bubble. I'm calling it now.

Actual AI doesn't understand the concept of the number four. It spits out grammatically correct nonsense. It is a very advanced pattern recognition system, and even its creators don't understand how it works.

It CAN be used for scanning resumes (that's a pattern recognition task), but human algorithms can do that, too, and probably better.

Training? Interviewing? Resolving employee problems? Setting policy? Meeting regulatory requirements? Talking to people? If any of the "AI HR companies" are claiming to have actually automated this using AI, I want to read about it. I'm finding marketing BS that often sounds like it's written by AI. I'm betting that what is being sold as AI is actually just outsourcing, with chat bots between customers and clients.

Are you willing to name the California company?

No. But it fucking happened. My wife is freaking out over what had happened to her job. Yes, they are doing all that shit. And laying off people. Mrs Quandry is lucky they found her a position. Because they fucking replaced her job with some bullsit they call AI because it sells. They would replace the fucking caregivers if they could. It is not fucking outsourcing. It's a motherfucking LLM chatbot that is expected to teach largely immigrants with limited computer skills how to do their fucking job.

Oh, and fuck you. I am not making this shit up. Fuck yourself with a rusty AI chainsaw. Asshole.

 

[Free]

We'll need an AI popcorn reaction. Soon.

 

[Noodles]

We'll need an AI popcorn reaction. Soon.

Write a python script that scrapes the forum,. then pushes it to an AI bot for sentiment analysis, then replies with the appropriate gif.

 

[Free]

Write a python script that scrapes the forum,. then pushes it to an AI bot for sentiment analysis, then replies with the appropriate gif.

Seems like a lot of work when I can just not do it.

 

[Veritable Quandry]

Seems like a lot of work when I can just not do it.

Just ask ChatGPT to write it.

 

[Anya Ristow]

I am not making this shit up.

I'm sorry. I know you're not making it up.

some bullsit they call AI because it sells.
It's a motherfucking LLM chatbot that is expected to teach largely immigrants with limited computer skills how to do their fucking job.

You are right, it is bullshit. LLM chatbots can't actually do that, and the patients will suffer. Very unfortunate for everyone.

I'm glad she found another position.

Are you willing to name the California company? I'd love to read about it.

If this is the cause of your ire, I did not mean this as "prove it." I meant it as "I want to read what they claim to be able to do." Because...

So now you have heard of a significant number of humans being replaced by AI.

Indeed, this is the first I've heard of anyone being replaced by AI.

I believe you.

 

[Veritable Quandry]

There will not be anything published. I have probably already posted more than the company wants said. Trade secrets/proprietary information.

 

[Argent Stonecutter]

The term is finally being applied appropriately.

I don't think it is. Large language models are no more AI than velcro is a rocket. It's a spinoff of AI research based on '50s-era models of neurons but there's no "intelligence" involved.

 

[Anya Ristow]

I don't think it is. Large language models are no more AI than velcro is a rocket. It's a spinoff of AI research based on '50s-era models of neurons but there's no "intelligence" involved.

Indeed, it doesn't actually understand anything. I've demonstrated that it doesn't understand the concept of the number four.

When I say it is finally being used appropriately I mean that people are agreeing on what it means (even if they are wrong), that procedural solutions are not AI, and you generally know what they're talking about.

I truly don't know how close GENERAL AI is. I suspect it isn't close at all. But LLMs smashed language parsing while I wasn't looking, so who knows?

 

[Casey Pelous]

... When I work with AI it feels like I am dealing with a high school student who has not read the text and is bullshitting their way through an essay.

This is the most spot-on description of the experience of "working" with "AI" I have seen.

 

[Argent Stonecutter]

When I say it is finally being used appropriately I mean that people are agreeing on what it means (even if they are wrong), that procedural solutions are not AI, and you generally know what they're talking about.

If they're wrong, they're wrong. Literally.

LLMs smashed language parsing while I wasn't looking, so who knows?

I don't know how much is language parsing and how much is pattern matching. The "taking a wolf across a river" example implies that it's more like a super-regex.

 

[Veritable Quandry]

It can't even really parse language since words are converted into tokens. Ask an LLM how many "r"s are in "strawberry." They may have patched it, but all of them had 2 instead of 3 for the longest time.

 

[Bartholomew Gallacher]

I know of nobody who has been replaced by AI. I have heard of nobody who has been replaced by AI. When I work with AI it feels like I am dealing with a high school student who has not read the text and is bullshitting their way through an essay.

Well I do remember that some months ago we had the story of a support hotline, which was there to gave moral support and guidance for obese people willing to reduce their weight.

They were replaced by a LLM AI chat bot.

The chat bot then told them the obvious "don't be lazy, eat less and move your body more", which was the reason why it was shut down again because the customers claimed that statements like this would make them eat more.