Mastodawn

Jerry 🦙💝🦙

I have this mental image of people on the bridge of a starship talking to the AI and instead of asking it meaningful things, they keep asking it to generate stupid cat memes and ways to prank their crewmates.

In my own work, I am seeing so much overreliance in under inspection of AI generated stuff that has consequential impact on their business. It feels like there’s going to be some sort of corporate Darwinism ahead based on companies misuse of LLM’s.

@jerry So the rats nibbling on the cables will be zombified? Or catified?

@jerry I've been using the "Star Trek" concepts to describe the AI panic. It really does seem like executives have globally decided to put their heads in the sand.

I really don't understand their mindset. When/if they get what they want, it's just going to harm them.

@jerry besides "enhance," the big screen on the Enterprise now also does "nudify" and a lot of other questionable things

Kyle Reddoch (CybersecKyle)5d ago

@jerry Corporate Darwinism is exactly the phrase.

Some companies are using AI like a junior analyst with supervision.

Others are using it like an unattended intern with domain admin, a company credit card, and zero change control.

Only one of those groups is making it back from the away mission. 😂

@jerry I think people are using it for too many things past their own intuitive "wait, that's not right" threshold so they're not noticing when it screws up. If they don't see it screw up, they don't have the inherent mistrust of the output they quite frankly need.

Chris Routh (He/Him)5d ago

Cybarbie 5d ago

Captain: Computer, cup of earl grey - hot!
Computer: Earl grey tea 1000 degrees
Captain: Aieee!!!
Computer: You're absolutely right! Warp core ejected.

swfinder 5d ago

@nf3xn @jerry You gotta appreciate that even the first Star Trek didn't trust the computer as far as it could throw it. Anything important is a person sitting at a desk doing it manually 😂

Misuse Case 5d ago

@jerry @oldredsubby From your mouth to God’s ear, Jerry.

Wendy Nather 5d ago

@jerry “Wesley, did you have the computer write this for you?”

DevSecOops 5d ago

@jerry I think infinitely more cat memes is the ultimate goal of generative AI. Job displacement is a regretable and IMHO naive consequence.

The next few years are going to be terrifying/really fun depending on your perspective. Anyone with responsibilities involving information security will likely fall into the prior category. Terrified!

Good luck!

I would like to once again refer to my prediction that at some point some CFO will just ask a bot to "Generate me next quarter's results".

All the talk I'm hearing from people over invested in the tech makes me think it's already happening.

Jerry 🦙💝🦙5d ago

@fennix allegedly AI is helping close the books in OpenAI and Anthropic and that is causing some consternation for their efforts to become publicly traded companies.

I think the nearer term general CFO action is to start clamping down on AI spending. That will happen and it might be the thing that finally kicks the leg out from this pyramid scheme

Michael T Babcock 5d ago

@jerry you know, a good #spoof series/movie/book about space travel in the era of bad-AI would be pretty funny actually.
Any writers out there willing to make us Babylon AI?

The Gibson in the desert 5d ago

@jerry So, Jerry, I think you know where I work... we've been using this tech well before it was a pop culture touchpoint. There are effective use cases there.

99% of what I see ain't that.

Jerry 🦙💝🦙5d ago

@thegibson It definitely has benefits for those who know how and where to use it. I’ve seen a lot of lazy nonsense in the past few weeks though.

The Gibson in the desert 5d ago

@jerry You know, we tried to market our capabilities as AI powered defense in 2017.

people were not ok with it.

So we stopped... NOW everyone asks what we're doing with AI, and I'm over here like... kinda did that a long time ago...

@jerry What I don’t understand is the lack of systems thinking. LLMs are one tool in a system, it’s not a system. It has inputs and outputs just like any other tool. There need to be evaluators to validate the output meets the desired outcome. I see so many uses that trust LLM output with either no or only minimal evaluation layers. Everything has failure modes. Enterprise systems account for failure as needed to meet the business outcomes. But too many people are throwing decades of enterprise systems design and operational practices out the window where LLMs are involved. I don’t get it.

K. Reid Wightman

@jerry related: https://simonwillison.net/2026/May/27/kyle-ferrana/#atom-everything

A quote from Kyle Ferrana

PICARD: Data, shields up DATA: Brilliant! Shields can reduce damage we sustain. Not immunity. Not hubris. Just prudence. It's not precaution—it's strategy. [camera shakes] WORF: HULL BREACHES ON NINE DECKS …

Simon Willison’s Weblog

Sterling 5d ago

@jerry
In my work experience, there was always "accountability". The upper managers always wanted to identify who is responsible.

AI will be the new "dog ate my homework".
That will not sit well, except with the lawyers.

@jerry
Oder Tee zu materialisieren
#anhalter

@jerry Will there be corporate Darwin awards?

even worse; statistical math infers a significant reduction in human brain plasticity with regular usage. no kidding.

(image Gimp-not AI)

Greg Bell 4d ago

@jerry survival of the fittest memes

@jerry Totally feeling this. AI produces so much output, it even put me back to mastodon because I think, here people still like to talk human-to-human.

James Tweedie 4d ago

@jerry
"I know I’ve made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal.
I’ve still got the greatest enthusiasm and confidence in the mission.
And I want to help you."

#HAL9000 / #Grok / #Claude / #ChatGPT / #Copilot / #MetaAI etc. 😉

Jessica Scott 4d ago

@jerry i could not agree more. It’s being used for majorly consequential things and no one is being held accountable for poor/deadly outcomes