Anybody who has worked in IT support in the trenches in enterprise IT will tell you there are some Excel power users who basically run the company, are macros wizards and actual ninjas.. about 0.1% of the workforce.
-
I’m really serious about this one btw. Companies have no measurable way of knowing what employees are doing with GenAI. They’re giving Claude Code out like it’s candy and just presuming everybody is an IT power user. They aren’t. They’re converting PDFs and vibe coding garden planning tools.
Copilot M365 has a fake dashboard showing how productive people are.. it has no actual data. It just shows people use it. It’s CIO porn for the CEO. Orgs are pissing money up a wall worldwide.
@GossiTheDog I'm not sure what an "IT power user" is anymore.
Those that chased the AI hype now don't seem to be able to do basic tasks without consulting one of the statistical make-believe machines. And those that aren't raking up token counts that look like phone numbers are apparently "unproductive"... despite some of those still having an understanding of the company systems.
-
I’m really serious about this one btw. Companies have no measurable way of knowing what employees are doing with GenAI. They’re giving Claude Code out like it’s candy and just presuming everybody is an IT power user. They aren’t. They’re converting PDFs and vibe coding garden planning tools.
Copilot M365 has a fake dashboard showing how productive people are.. it has no actual data. It just shows people use it. It’s CIO porn for the CEO. Orgs are pissing money up a wall worldwide.
> They’re giving Claude Code out like it’s candy and just presuming everybody is an IT power user.
It's worse at some places. There "let's use more AI is a target" and the number burned tokens basically becomes a KPI.
An since there is a very real danger that one's AI use (or lack thereof) become a topic in a future performance evaluation, I have every respect for people who say: They want us to burn tokens, so let's burn tokens.
Best idea actually is, IMHO, to ask every inane question of the LLM, then NOT look at the results to avoid getting dumber and just work out your report / solution whatever by traditional means. Have you used AI? But certainly, Sir. The results had to be heavily edited of course...
-
@GossiTheDog I'm not sure what an "IT power user" is anymore.
Those that chased the AI hype now don't seem to be able to do basic tasks without consulting one of the statistical make-believe machines. And those that aren't raking up token counts that look like phone numbers are apparently "unproductive"... despite some of those still having an understanding of the company systems.
If it's IT as it's set up today, we can assume that there is no overlap with the set of power users. And if it's about people who use IT provided tools: The same applies.
The power users are those who use (against heavy resistance) tools like emacs, Jupiter, R or simple plain programming languages. These users are not supported by IT at every step.
-
I’m really serious about this one btw. Companies have no measurable way of knowing what employees are doing with GenAI. They’re giving Claude Code out like it’s candy and just presuming everybody is an IT power user. They aren’t. They’re converting PDFs and vibe coding garden planning tools.
Copilot M365 has a fake dashboard showing how productive people are.. it has no actual data. It just shows people use it. It’s CIO porn for the CEO. Orgs are pissing money up a wall worldwide.
This isn’t, btw, a Kevin Doesn’t Use AI rant. I use GenAI extensively for various testing things. I had access to Mythos before almost everybody. I pay for Copilot Pro+. I pay for Gemini. I AI generate terrible songs. I vibecode security scanners. My personal spend exceeds £500 this month to date.
Do I think you should make every employee depend on these third party GenAI tools for their job? No. Its ridiculous. You’re also *locked in* to costs you can’t afford which *will rise*. It’s a cliff.
-
This isn’t, btw, a Kevin Doesn’t Use AI rant. I use GenAI extensively for various testing things. I had access to Mythos before almost everybody. I pay for Copilot Pro+. I pay for Gemini. I AI generate terrible songs. I vibecode security scanners. My personal spend exceeds £500 this month to date.
Do I think you should make every employee depend on these third party GenAI tools for their job? No. Its ridiculous. You’re also *locked in* to costs you can’t afford which *will rise*. It’s a cliff.
@GossiTheDog absolutely 100% with you on this
-
This isn’t, btw, a Kevin Doesn’t Use AI rant. I use GenAI extensively for various testing things. I had access to Mythos before almost everybody. I pay for Copilot Pro+. I pay for Gemini. I AI generate terrible songs. I vibecode security scanners. My personal spend exceeds £500 this month to date.
Do I think you should make every employee depend on these third party GenAI tools for their job? No. Its ridiculous. You’re also *locked in* to costs you can’t afford which *will rise*. It’s a cliff.
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
-
This isn’t, btw, a Kevin Doesn’t Use AI rant. I use GenAI extensively for various testing things. I had access to Mythos before almost everybody. I pay for Copilot Pro+. I pay for Gemini. I AI generate terrible songs. I vibecode security scanners. My personal spend exceeds £500 this month to date.
Do I think you should make every employee depend on these third party GenAI tools for their job? No. Its ridiculous. You’re also *locked in* to costs you can’t afford which *will rise*. It’s a cliff.
@GossiTheDog This is why I very seldomly use cloud AI and on purpose try to do everything with a local LLM setup. We know the subsidized cloud plans are going to come crumbling down so it's better to not have becomed dependent on them in your workflow.
-
@nils_ballmann it’s happening worldwide in every sector. It’s just incredible, and very, very funny.
@GossiTheDog From a (rather privileged) perspective that one's job and life is (rather) unaffected from the AI bubble bursting I can imagine it to be (strained) funny.
From the perspective of being responsible to pioneer a thing that is not unlikely to be cut just to save the company... I just want it to be over with.
I want the bubble to pop and start enduring the ensuing chaos. It would finally get rid of the uncertainty.
-
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
@GossiTheDog honestly, isn't most of the stuff IT teams do nowadays extremely wasteful?
I look at CI pipelines and feel the need to scream.
Upgrade your OS image with hundreds of packages on every push,
Build all layers of your container every time...
Then migrate to the next git service and CI/CD framework every year, complete rewrites.
How many bloody Artifactory mirrors does a company need?!!!Etc. etc. These kids should start with a C64 or ZX80 before let loose on this hot garbage
-
RE: https://mastodon.social/@404mediaco/116908074107231828
Anybody who has worked in IT support in the trenches in enterprise IT will tell you there are some Excel power users who basically run the company, are macros wizards and actual ninjas.. about 0.1% of the workforce. About 99% of people can’t align a table in Word.
Giving the 99% of people tools which cost $$$ per user a month and letting them do anything is like giving a child a car, and being surprised when they ram the car into a wall three days later and cost $10k after achieving nothing.
@GossiTheDog this can also be observed from actual car usage. I watched a dumbass who shouldn't be driving render a roundabout inoperable because they came to a stop IN the roundabout waiting for someone to enter it.
In a neighborhood where the average home price is well into the 7 figures.
-
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
Business are handing money to AI companies and getting absolutely nothing useful back for 99% of that money.
It would be more productive to burn actual cash money in their heating systems or even buy fixed assets such as Gold or DIMM modules.
Or throw the money at FLOSS projects so they are better funded.
-
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
@GossiTheDog For awhile, I had to report how much productivity AI added to my work. Now, I can no more quantify it that I can quantify the benefit of using a good IDE, adding extra RAM, or getting fresh air. Of course, negative numbers were not allowed.
-
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
@GossiTheDog This is stupid, of course - but, as they say, show me the incentive and I'll show you the outcome.
Still, I don't quite understand the "cost" thing. If used properly, AI should reduce cost. By using Claude, I was able to do in 3 months what I previously couldn't do in 5 years. This is on a Pro subscription (the cheapest one) that costs something like 180 euros per year.
If we had hired a professional programmer for the same work, their *monthly* salary would have been more than 10 times higher - and they would have become annoyed with me and left after 2 months anyway.
-
I’m really serious about this one btw. Companies have no measurable way of knowing what employees are doing with GenAI. They’re giving Claude Code out like it’s candy and just presuming everybody is an IT power user. They aren’t. They’re converting PDFs and vibe coding garden planning tools.
Copilot M365 has a fake dashboard showing how productive people are.. it has no actual data. It just shows people use it. It’s CIO porn for the CEO. Orgs are pissing money up a wall worldwide.
@GossiTheDog I am no general opponent of AI use. But as every tool it takes a lot of learning.
The shallow learning curve at the beginning fools users and managers alike.
For long term reliable and reproducible success you need a lot of structure.
Developing that and enforcing it is a real challenge for most users.
-
@GossiTheDog This is stupid, of course - but, as they say, show me the incentive and I'll show you the outcome.
Still, I don't quite understand the "cost" thing. If used properly, AI should reduce cost. By using Claude, I was able to do in 3 months what I previously couldn't do in 5 years. This is on a Pro subscription (the cheapest one) that costs something like 180 euros per year.
If we had hired a professional programmer for the same work, their *monthly* salary would have been more than 10 times higher - and they would have become annoyed with me and left after 2 months anyway.
@bontchev @GossiTheDog Why would your hypothetical employee have become annoyed with you?
-
@GossiTheDog This is stupid, of course - but, as they say, show me the incentive and I'll show you the outcome.
Still, I don't quite understand the "cost" thing. If used properly, AI should reduce cost. By using Claude, I was able to do in 3 months what I previously couldn't do in 5 years. This is on a Pro subscription (the cheapest one) that costs something like 180 euros per year.
If we had hired a professional programmer for the same work, their *monthly* salary would have been more than 10 times higher - and they would have become annoyed with me and left after 2 months anyway.
@bontchev @GossiTheDog Are you paying the full price, which covers all cost or are you using AI burning VC capital?
Edit: Stupid question, I overlooked that you're paying 180 €. So burning VC capital.
-
@GossiTheDog I am no general opponent of AI use. But as every tool it takes a lot of learning.
The shallow learning curve at the beginning fools users and managers alike.
For long term reliable and reproducible success you need a lot of structure.
Developing that and enforcing it is a real challenge for most users.
> I am no general opponent of AI use.
I am, though.
> Developing that and enforcing it is a real challenge for most users.
Is it even worth the trouble or feasible? I'd like to remind people that this is the same group of users where it was too much hassle to learn "the command line" or even how a filesystem works. It all needed to be point & click "discoverable" interface (never mind that this promise was never fulfilled).
The point is discoverability, though: An interface where function is (visually) recognizable.
I wonder how (chat based) AI --- where no feature is actually _visible_ --- will work for this user group.
Yes, I think it has a use, though: As a normal, only smart feature. Like cropping subjects from a picture as in Apple Photo (on iPad, YMMV): It's just a very smart lasso, but the concept is the same as of any other cropping tool: There is a line and the image is cropped along the line (if the user wants it).
But it doesn't have a use a smartass ghost in the machine which does magic things or nags or bosses around the user.
The average user won't "develop" a "structure", though: We (the software industry) have weaned them from doing so successfully now for almost 3 decades. The capability is gone. The promise of achievement without effort has stayed ...
-
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
@GossiTheDog Show me a metric and I'll show you a game.
-
I’m really serious about this one btw. Companies have no measurable way of knowing what employees are doing with GenAI. They’re giving Claude Code out like it’s candy and just presuming everybody is an IT power user. They aren’t. They’re converting PDFs and vibe coding garden planning tools.
Copilot M365 has a fake dashboard showing how productive people are.. it has no actual data. It just shows people use it. It’s CIO porn for the CEO. Orgs are pissing money up a wall worldwide.
@GossiTheDog i've found it to be quite difficult to use the various security copilots for their (supposedly) intended uses without simply swapping previously automated processes for them, which is insane but also what one rather supposes to be the point.
i have however also used generic corporate copilot to do arbitrarily silly things like deobfuscate malware so anybody snooping on my prompts in the name of metrics or other glory may be in for a surprise
-
In an era where companies need to become more efficient and diverse they’ve basically picked the least efficient way to do it, with the biggest risks and highest costs - because everybody else is doing it.
I know somebody at one of the big 4 who has written something in Claude that prompts Claude each twenty minutes for a question, then feeds Claude’s question back into Claude to use their tokens - because token usage is factored into employee evaluations. What are we even doing.
@GossiTheDog I wish I could find the original to attribute it properly but this fits
