I work in government policy, my department has a team devoted to tech and "transformation," and of course AI is the latest darling.
They did a demo recently to show what it can do. You can enter a prompt and it will write a policy!
But as I told my boss recently, do you know how hard it is to write the actual policy? I can fart one out in an hour. Writing the thing is not the hard part!
Most of policy analysis is figuring out why people aren't implementing policy that's already been written (does it not make sense, is it not possible with current caseloads, are people acting in bad faith, etc). AI won't help with any of that.
The problem is it will probably take ten years of trying to make it work and laying people off before the people who make these decisions give up. I'm sure some fucker in a state or municipal governement somewhere has already drafted plans for getting rid of the people who think about government problems for a living on the belief that AI can do this instead.
Is that belief sincere? I honestly don't know. I'm not sure it makes a difference, although I'm with you, Vince: the people that piss me off the most are the ones that cheerlead how we're all going to get robofucked. They want AI to be Data from Star Trek, instead it's a Sybian duct taped to a mannequin.
I write corporate communications for a living. I want to have an open mind. Therefore, on a few occasions, I have asked ChatGPT to provide me with research sources that support a certain claim. It generates a list but when I go to the direct source there's nothing there.
So I ask ChatGPT for the links to the source it provided. Every time I do this the links lead to nowhere. Not even a 404 error. Just a link that does nothing when you click it.
I think the most astute assessment some have made is that AI is "eager to please." Like a hopeful, idealistic, young intern, it doesn't dare tell you "I don't know." Rather, it makes up shit on the spot to save face.
Of course, AI evangelists have wasted no time responding to this legitimate criticism by reminding people that they "need to get better at writing their prompts." That's right, they think the problem is that we've asked the wrong question, not that the answers come up empty.
No matter how many times and ways I think about all of this I keep arriving at the same conclusion, we all just need to go outside more.
I work for a company which constantly promotes it's green/Net Zero values, but inevitably has latched on to AI like white on rice. Is it just a plagarism machine built with slave labour with the carbon footprint of a fleet of private jets? Sure. But can't get in the way of #growth!
I’m a designer/illustrator and I have had multiple AI “artists” explain to me how it takes just as much skill and practice to write prompts perfectly to get the art you want out of AI, which no, but even if it were true, why? What would be the point of training up an entire other whole person to exert equal effort to replace an existing job just to include a robot middleman? Because tech? The future is so fucking stupid lol
AI reminds me of the CBD boom. The believers were pushing a wild agenda. Do you need energy? CBD. Do you need to relax? CBD. Joints hurt? CBD. Hair loss? CBD. At a certain point it didn't make any sense. Anyway, I've got to go. I'm meeting Van Jones down at the park so we can pop on our Google Glasses and ignore calls from our families.
Another thing it reminds me of is the metaverse. You can bet that if Mark Zuckerberg hadn't renamed his company Meta just a few years ago, he would be renaming it fAIcebook or something like that now. Hopefully AI will ultimately flop too.
I'm glad I work for a midsized company where AI wouldn't be scalable for anything we do. That said, any time it is brought up in a manager meeting I do my best to shoot it down immediately.
There is, however, a big push in the high tech manufacturing industry for adopting cobots (goofy-ass name) for automation of some assembly tasks. I'm the guy who usually does the ROI for our production improvement capital projects, and I hate to say it, but if they work they are a solid investment with a quick payback period. The 'if they work' statement is a big 'if' though.
While I feel conflicted about replacing human labor with automation, the truth is the manufacturing industry in the US (and in the northeastern US especially) is in a labor deficit. We have a very hard time filling positions to keep up with our growth despite offering well above market average wages for hourly positions. It's not an industry that is attractive to young people and a lot of the workforce is at or near retirement age. There is a lot of blame to go around for why this is (dismantling unions and decreasing wages for the past 40 years being the primary drivers) but if the labor market doesn't improve you are going to see much higher adoption of automation in the next 10 to 15 years.
I'm a high school teacher. We have all these safeguards put in place to prevent our students from using AI (which isn't tough to detect, it's usually off-topic dog shit). And yet every time we have a staff meeting they show us some cool new AI we can use to help create quizzes and lessons. These are also dog shit. I honestly can't blame students for turning in terrible AI assignments when all their teachers are creating terrible AI curriculums. There's even an AI that writes letters of recommendation, which really depresses me. We're creating a future where AI takes all the jobs from kids who were never educated in the first place so that everyone can just become the human batteries from The Matrix that help fuel the AI machines.
I work in government policy, my department has a team devoted to tech and "transformation," and of course AI is the latest darling.
They did a demo recently to show what it can do. You can enter a prompt and it will write a policy!
But as I told my boss recently, do you know how hard it is to write the actual policy? I can fart one out in an hour. Writing the thing is not the hard part!
Most of policy analysis is figuring out why people aren't implementing policy that's already been written (does it not make sense, is it not possible with current caseloads, are people acting in bad faith, etc). AI won't help with any of that.
The problem is it will probably take ten years of trying to make it work and laying people off before the people who make these decisions give up. I'm sure some fucker in a state or municipal governement somewhere has already drafted plans for getting rid of the people who think about government problems for a living on the belief that AI can do this instead.
Is that belief sincere? I honestly don't know. I'm not sure it makes a difference, although I'm with you, Vince: the people that piss me off the most are the ones that cheerlead how we're all going to get robofucked. They want AI to be Data from Star Trek, instead it's a Sybian duct taped to a mannequin.
How much do you want for the Sybian mannequin?
I write corporate communications for a living. I want to have an open mind. Therefore, on a few occasions, I have asked ChatGPT to provide me with research sources that support a certain claim. It generates a list but when I go to the direct source there's nothing there.
So I ask ChatGPT for the links to the source it provided. Every time I do this the links lead to nowhere. Not even a 404 error. Just a link that does nothing when you click it.
I think the most astute assessment some have made is that AI is "eager to please." Like a hopeful, idealistic, young intern, it doesn't dare tell you "I don't know." Rather, it makes up shit on the spot to save face.
Of course, AI evangelists have wasted no time responding to this legitimate criticism by reminding people that they "need to get better at writing their prompts." That's right, they think the problem is that we've asked the wrong question, not that the answers come up empty.
No matter how many times and ways I think about all of this I keep arriving at the same conclusion, we all just need to go outside more.
I work for a company which constantly promotes it's green/Net Zero values, but inevitably has latched on to AI like white on rice. Is it just a plagarism machine built with slave labour with the carbon footprint of a fleet of private jets? Sure. But can't get in the way of #growth!
Also, Ashton Kutcher continues to be the living embodiment of his first name. As 30 Rock put it "there will never be a president Ashton"
I’m a designer/illustrator and I have had multiple AI “artists” explain to me how it takes just as much skill and practice to write prompts perfectly to get the art you want out of AI, which no, but even if it were true, why? What would be the point of training up an entire other whole person to exert equal effort to replace an existing job just to include a robot middleman? Because tech? The future is so fucking stupid lol
AI reminds me of the CBD boom. The believers were pushing a wild agenda. Do you need energy? CBD. Do you need to relax? CBD. Joints hurt? CBD. Hair loss? CBD. At a certain point it didn't make any sense. Anyway, I've got to go. I'm meeting Van Jones down at the park so we can pop on our Google Glasses and ignore calls from our families.
Another thing it reminds me of is the metaverse. You can bet that if Mark Zuckerberg hadn't renamed his company Meta just a few years ago, he would be renaming it fAIcebook or something like that now. Hopefully AI will ultimately flop too.
I'm glad I work for a midsized company where AI wouldn't be scalable for anything we do. That said, any time it is brought up in a manager meeting I do my best to shoot it down immediately.
There is, however, a big push in the high tech manufacturing industry for adopting cobots (goofy-ass name) for automation of some assembly tasks. I'm the guy who usually does the ROI for our production improvement capital projects, and I hate to say it, but if they work they are a solid investment with a quick payback period. The 'if they work' statement is a big 'if' though.
While I feel conflicted about replacing human labor with automation, the truth is the manufacturing industry in the US (and in the northeastern US especially) is in a labor deficit. We have a very hard time filling positions to keep up with our growth despite offering well above market average wages for hourly positions. It's not an industry that is attractive to young people and a lot of the workforce is at or near retirement age. There is a lot of blame to go around for why this is (dismantling unions and decreasing wages for the past 40 years being the primary drivers) but if the labor market doesn't improve you are going to see much higher adoption of automation in the next 10 to 15 years.
I'm a high school teacher. We have all these safeguards put in place to prevent our students from using AI (which isn't tough to detect, it's usually off-topic dog shit). And yet every time we have a staff meeting they show us some cool new AI we can use to help create quizzes and lessons. These are also dog shit. I honestly can't blame students for turning in terrible AI assignments when all their teachers are creating terrible AI curriculums. There's even an AI that writes letters of recommendation, which really depresses me. We're creating a future where AI takes all the jobs from kids who were never educated in the first place so that everyone can just become the human batteries from The Matrix that help fuel the AI machines.
Great piece. Isn't it "short sighted" though?
I THOUGHT I ALREADY FIXED THAT, GREG. But apparently I'm an even bigger idiot than I thought because I typed it wrong twice.
After watching the US Open yesterday I started thinking I had it wrong the whole time.