Disclaimer: I don’t have any understanding when it comes to finance/economics other than what I’ve experienced/read. However, it does seem hard to argue the fact that economics/finance are so divorced from reality that they ought to be viewed the same as the national debt.
I’m sure there are others in the same boat, but I’m still having trouble getting past the absurdity and ominous nature of everything happening in the corporate world right now. Almost every single conversation is boiled down to “just ask Claude” and it’s being used to slow down existing workflows to a comical degree. Who knows maybe this will self-correct over time, but it’s certainly a choice at the moment.
I’m not trying to be a doomer I’m really not, but these people live in a different reality than you and I. I used to listen to Ed Zitron’s podcast and tuned in to any number of his appearances on other shows to sort of shield myself from all the noise that encompasses this age of tech, but I’m slowly coming to the realization that the economics/financial picture of the situation seems entirely inconsequential. I’m hoping this is just my lack of knowledge when it comes to finance/econ, but fuck if I know. Maybe this post will make me look like a dumbass in a year or so (I hope it does).
I feel like all of us in corporate are all assets in a bet from the most evil people on the planet to see if they will be able to replace huge swaths of the population with a “good enough” solution that will slowly be de-valued and phased out of the workforce akin to IT jobs in the early/mid 2010s.
This shit really blows and I feel like LLM companies are going to come out on top, primarily because Claude has proven that it can generate passable output at lightning speed and that is more than enough to please C suites.
One of the higher ups pointed ChatGPT at our GitHub and asked it to find things for us to do. They then sent us the resulting four pages of incoherent consultant speak it extruded and asked us to address the issues.
I’m pretty sure one of the reasons LLMs are catching on with bosses is because they’re used to throwing gobs of money at real people to produce this sort of work-adjacent bullshit and so the ability to automate it is an actual innovation in their minds.
The sequel to Office Space in which Lumbergh fired all of his employees, and then uses his signature passive aggressive tone to ask ChatGPT to give him stuff like TPS reports. Milton gets ready to burn it all down for a second time.
Cory Doctorow used the metaphor of a toddler sitting in the back seat with a toy steering wheel to suggest we let billionaires play with LLMs while quietly airgapping them away from anything that society actually requires to function. If management could be contented with a sycophantic algorithm churning out nonsense reports while someone who drew the short straw occasionally goes in and reassures them that their instructions are being followed, that could be the best solution.
That makes me think of the “are ya winning, son?” meme, but the son is a CEO who contributes nothing to society, and the dad periodically checks up on him to give encouragement and to see if the latest LLM is still keeping him entertained and safely away from society.
They probably wouldn’t even notice if we were sneaky enough about it!
Executive at my company said something along the lines of “AI isn’t leaving, it’s a massive productivity boost” and then in the same breath said mfs needed to stop causing production outages and that we were on a change freeze for the next month.
I worked at a national telecom company where someone would fuck something up somewhere, and in response we’d have a nation wide, company wide, “alright pencils down everyone” week or two where nobody was allowed to touch anything.
Change management just spent time telling everyone to stop working.
More and more of my customers (plumbing) are approaching me during diagnostic procedures and telling me “Clob says this should only take XYZ amount of minutes and you haven’t found ABC intermittent currently nonexistent issue because you are wrong”.
“Clob says it’s not the shower cartridge it’s the flux capacitor so would you check that instead?”
So I check the flux capacitor and surprise surprise, it’s fluxing as it should be. After exhausting their AI inspired options, I go back to my original plan and it’s buttoned up inside thirty minutes. Often, they ask for reduced labor costs because once I found it it only took me thirty minutes to fix, never mind the hour I spent placating a tight ass homeowner.
Or I show up and they tell me “XYZ is broken here’s a replacement part on the counter I asked AI so just change that please” it’s never the right replacement part and Clob is never right. More often than not I have to leave to pick up the right part and they get upset they’re charged for my windshield time.
AI shit sucks and folks who use it suck.
What’s up with homeowners and trying to be as cheap as humanly possible? Like my dad got a quote for our roof for 6 grand, so he told them to beat sand and asked his friend to do it for the price of parts. I sure hope he knows what he’s doing because his friend isn’t a roofer, he’s just a general handyman.
Similar to car mechanics, a lot of these private businesses will just try to scam you. If it comes down to trust, well, I know I can trust my friend, but who knows if that random roofing company aren’t just going to fuck me out of several thousand dollars?
Suburbanites and their libertarian brainworms.
Also the industry being filled with people taking advantage of presumed wealthy people (cuz, you know, house) who don’t know better and not being able to tell a fair price. Got quoted 800usd to replace a water main shutoff valve stem last year, as an example.
That’s correct. The industry is also fucking FULL of private equity plumbing companies that nickel and dime for everything. An example; as a routine part of chasing a sewer clog sometimes I need to pull a commode. I will typically just re-use the commode supply line unless it is leaking, visibly damaged etc. Lot of places will not pull a commode without requiring you to purchase a new commode supply line due to “liability”, they claim it might rupture because it’s reused and it must be replaced instead. That supply line costs me about three bucks wholesale and then places like Roto-Rooter will charge you forty bucks for the supply line and fifty bucks for labor.
The shop I work for gets a bit exorbitant even then. Example sewer call:
$145 trip charge just for me to darken your door
$175 per hour for me (TWO HOUR MINIMUM!)
$85 per hour for my greenhorn
$100 equipment charge
$50 gas
$25 consumables charge (gloves rags shoe booties etc)That’s how you end up with an $25 dollar brass valve costing 800 to install
I will however, find any and every way possible to stick you with as large a bill as I can legally manage if you are a land lord/property manager etc. I will lie to landlords about the tenants responsibility for clogs/breaks/negligence. I absolutely enjoy that part, comrade. It’s never the tenants fault, even if it is.
It’s never the tenants fault, even if it is.
Based.
What’s up with homeowners and trying to be as cheap as humanly possible?
My landlady has “cheap as humanly possible” mindset. For any kind of repair work - she hires the cheapest guys possible and then she gets pissed off when they do a shitty job. Luckily for me when I had a rain leak in my living room - she spent a bit more so the guys did a good job. But I’d spy a look at them and they’d be on the very dangerous “do not step” part of the ladder. I half-expected one of the guys to fall off and break his neck.
Tbf I’m a pretty good plumber and I’ve used the “not a step” part of the ladder. I’ve also used the “do not climb” side of the ladder b/c sometimes it’s Friday 4:45pm so I’m not making another trip for different ladder.
The guy at the top was leaning too. It wouldn’t take much for the ladder to topple over. I don’t know how much time they spent in that perilous position but they did it early in the day and the job took three days.
Oh no. Thats actually a really good way to end up paraplegic first thing in the morning. I’ve used the top step on occasion to snag a hand tool or something left on a suspended ceiling or on top of a wall, or climbed the wrong side to retrieve a tape measure left on top of the ladder, but I won’t actually work up there. I’m the OSHA competent person where I work so I can’t bring doing dumb stuff like that too often.
Earlier today when I was commenting on AI I said that people love convenience more than common sense. I never thought of convenience as a siren song but it certainly is.
The trick to being a cheapskate is knowing when not to cheap out on stuff.
If it’s safety related or could destroy something, don’t cheap out on it.
I like the way you write
lol thanks
I actually know someone like this who had a medical issue, asked ChatGPT about it, concluded it was the solution to their ailment, and now swears by it
Same. A guy I know was all like “chatgpt fixed my lifelong back problems”. Then proceeded to tell me what exercises it told him to do and turns out most were the same ones I told him to do like ten years ago when I was his trainer. But he never really listened to me, because I am woman and obviously have no idea what I’m talking about.
Anyway, the “ai” told him to overdo it and do way too monotonous stuff so he did, and now his back is worse than it’s ever been.
I hadn’t even considered the effect prolific use of LLMs for literally everything would have on mansplaining. I mean why trust a woman on any subject you don’t know anything about when you can just ask ChatGPT instead.
I’ve thought about the way these are mansplaining machines a lot and how there would be space for some great feminist analysis of expertise, the way these things are so bro-coded. The impact that would have on society. It probably exists, but haven’t seen it.
It was wild realizing that a man took basically the same advice I had tried to give, but without the nuance, which would be needed to not fuck themselves up with it, when it came from a machine. This person is a tech bro who tends to think humans make mistakes, machines don’t. Big believer in robotic driving.
I heard a person saying their AI agent did this. I still don’t get what the fuck AI agents are. Is it making up a guy for the yes-man machine to pretend to be?
An “ai agent” is when you give the hallucination machine direct access to tools so they can do stuff autonomously. Like if you were to set it up so you can say “buy some socks” and it just goes on amazon and finds whatever sounds like socks and buys it for you. and if it decides that you need $8000 worth of socks, so be it
We had things that could do that well before these stupid chatbots. And they’d ask for confirmation before doing the thing. I guess people are just obsessed with the sci fi idea of “computer, resolve this mundane errand”.
ⓘ This user is suspected of being a bear. Please report any suspawcious behaviour.
Is there a way to make the doo doo and pee pee flow into their house instead of water? Because you should have done that.
From every economic measure ive been seeing, most things are telling that a recesion is coming, and soon, the actual debt of AI is much higher than we are being told as they use shell companies to hide the debt and lease their data centers to them. AI is not making money, no return on investment, at most its goal is to replace workers, its not happening anytime soon at the rate they want, yet the stocks price is exactly dependant on that speculation. CAPE is at 1999 and 2007 points, just before the previous crisis. US debt is unpayable, they paid this level of debt once before only because ww2 made the US the only remaining intact economy and they owned the world. Now? They have no industry, still on trade deficit, dependant on the military industrial complex and venture capital money laundering. The petrodollar is breaking with the war on Iran, absolutelly ridiculous fumble, shooting themselves in the stomach. Bond market is starting to break too, goverments dont trust US to pay its debt so they sell US bonds, so interests are higher, so debt repayments are higher, leaving the government with even greater deficit. They are issuing new debt to pay old debt in a doom spiral.
Material conditions will worsen considerably, I hope US people can wake up and gain some class conciousness.
US hard power allowed soft power to work, now both hard power is failing as we see in Iran, so the already weak soft power is crumbling as they cant keep up their world order.
The only choice is to ask China for help lmao. Century of US humiliation is on its way.
Im an armchair marxist economist at best, if you know more than me pls tell me so I can learn, thanks.
We’re driving head first into a wall, but we’re doing it aT tHe SpEeD oF AI! Isn’t that exciting?
I had a conversation with a long time developer today about this. He basically said his job has moved from “write reviewable code so it is easy to understand and maintain” to “spend all day reading a novel of code that technically works but would be completely unmaintainable if a human had to.”
Like technically the code works and so the “writing the code” part goes faster but then it takes the same amount of time to review, cleanup, refactor, fix, etc as it would have to just write the code right in the first place.
It’s basically shifting the problem space. The machine can solve the problem for you well enough but it takes a real expert to understand what it did and if it solved the problem well enough within the context of everything else going on too.
So it’s not really saving time, and therefore probably isn’t saving money in the long run.
Once the real costs of these machines starts coming down the pipe, I think most teams will realize that at least for doing complex tasks like writing code, it’s cheaper to just have a person do it.
Don’t get me wrong the machines can often write really impressive and correct code, but they are only doing it “faster” if you ignore the context and they’re only doing it cheaper because they’re being subsidized to do it cheaper. Once that subsidy goes away I think we’ll see a lot less of this stuff.
I do think it’s never going away though, I just don’t see a continued world where “just use AI” continues to be the solution to every problem. Especially not complex ones.
Wall Street opened a gigantic credit account for US companies to take on massive technical debt
Like technically the code works and so the “writing the code” part goes faster but then it takes the same amount of time to review, cleanup, refactor, fix, etc as it would have to just write the code right in the first place.
This has been my experience as well.
Charitably, there have been moments where I have used an LLM to help debug a problem. I was using Zig and trying to decompress and deflate a ZIP archive in memory and the documentation for the APIs in Zig leaves much to be desired. I was able to post a stack trace of the crash and slowly determine a solution but at no point was it where I just let the LLM write the code. It was much more “look at the generated code, determine the change that needs to be made, and try that” with some instances being where the suggestion was completely wrong. BUT, I eventually came to the solution. I would not have been able to figure it out on my own, but it absolutely was not the case where it’s letting the LLM do it and I just review the code.
A similar issue I had was with a crash with mariadb. New version of mariadb gets released and suddenly in development we are getting crashes right at the start of our test runs. I copy out the stack trace from mariadb and paste it into Copilot, and get a search result with some prodding, back to a bug when grant tables are disabled. It was very helpful to understand why, and our fix was to pin our mariadb version at a previous stable version
So, really it’s only useful as a search engine. I would not use it to just generate code.
What staggers me is how stupid/unimaginative people whose job it is to “solve business problems” are.
Like, I’m treated like a genius when I make a little Excel template to keep my boomer coworker from manually updating a sheet to track her higher value clients’ payments and orders. I guess I just get paid to solve little puzzles, idk I’ll enjoy it for what it is.
i made a pivot table the other day for organizing some response data to a survey and demonstrated how it works/how to use it in a meeting. people acted like conjured a portal to the future.
people acted like conjured a portal to the future.
I have this experience just with lookup tables and conditional formatting.
I never thought I’d see a thread this relatable about this exact topic
Sumifs statements are incomprehensible magic to the average office worker. Keeps people like us looking great though so I’m ok with it lol.
My job’s been trying to integrate Claude into our workflows and I was asked to try out the Excel addon. I have a decently complex sheet that I use to do an inventory audit. As a test, I pasted the pure data from that into a fresh spreadsheet, then gave it instructions to do the audit. Not only did it not give me what I asked for, it made like two or three extra tabs to find values that I only needed one formula to get. No clue how many tokens I would have had to burn through to get it into a usable shape. Oh and now Excel is even more sluggish regardless of if I’m using Claude. Definitely worth the trillions we dump into this bullshit, great job everyone
LLMs, for some reason, seem to perform worse when i get them to do something I’m knowledgeable of.
Weird
You’re just using the wrong one. ChatPPB is actually good
Y’all need to have it take it for you
Yeah, they’re pretty bad at the stuff I understand well but for everything else they’re incredible.
My brain functions!
I got the shitty Copilot “upgrade” for Excel and it always tries to autofill the worlds most pointless xlookups when I’m writing a formula. I tell it to make the formatting “professional” when it’s some kind of doohickey I’ll be handing off to another person (I rarely care about blinging out shit in Excel lol).
I felt like a dummy when I barely got my undergrad in math ages ago, but damn do I feel like the smartest computer toucher in the room too often.
Every time they treat you like a genius in front of the boss immediately ask for a raise.
OpenAI and Anthropic are now propped up by the US government to develop cyberweapons, bioweapons, etc. LLMs are actually useful in those cases because rapid, guided trial and error is a viable strategy. They only have to be right once, and it’s easily verifiable. So it’s definitely not going away.
I’m interested to see what happens in the general economy though, because I strongly believe LLMs are a net negative for productivity over the long term in most real work. In a way, the paying users are subsidizing it for the government.
generate passable output at lightning speed
It seems to me that’s the AI grift in a nutshell. People love convenience more than common sense.
