TechTakes

God save me from arguing with HPMoR readers.

personal vent: at my job yesterday i had to come up with a few fake book titles/author combinations for a project. a fun little task and opportunity to hide some cheeky easter eggs. so, i came up with a few and then asked my coworkers to share in the fun. one of them though just couldn't come up with anything at all, and eventually just resorted to "asking chat gpt".

mind you, i work a creative job, and so do my coworkers. this is a minor thing i guess, but it just made me very sad. how could you just outsource your creative joy to some mindless word salad machine?

A full timeline on the RubyGems takeover has been put together - looks like the entire situation's been caused by pressure from Shopify.

Techdirt: AI Slop Startup To Flood The Internet With Thousands Of AI Slop Podcasts, Calls Critics Of AI Slop ‘Luddites’

They say "Luddites" as if it's a bad thing...

Harvard Business Review: AI-Generated “Workslop” Is Destroying Productivity

[...] Employees are using AI tools to create low-effort, passable looking work that ends up creating more work for their coworkers. On social media, which is increasingly clogged with low-quality AI-generated posts, this content is often referred to as “AI slop.” In the context of work, we refer to this phenomenon as “workslop.” We define workslop as AI generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task.

They put 'environmental impact of AI' on the front of the student newspaper (below the fold, but still), then you flip and see this

kinda feeling two steps forward, three steps back rn on top of all the other drama on campus

This scream into the void has been on my mind for a while: Apparently I work for an AI company now.

Kinda.

When I had the interviews with my now-employer at the beginning of the year, they were an open-source cybersecurity startup. Everything sounded great, we got along, signed the contract. I took a long vacation before starting the position, and when I got back, I was... amused? bewildered? to find that a), we are no longer open source; and b), we have pivoted, hard, towards AI.

Luckily, I still get to work 100% of the time on the core (cybersecurity) product (which is actually a really good and useful thing, sorry, not going to be more specific), it's just that part of the dev team, as well as all of marketing and sales, now work on building and selling an AI product built on top of that.

At least it's not a wrapper around ChatGPT, and does offer something kinda new and actually beneficial, but still, it's an LLM product.

Now, for the actual scream-into-the-void: Once a month, in a company-wide meeting, I have to observe how people praise LLMs to the the moon, attribute nonsense or downright bugs to something akin to proto-sentience, and give absurd estimates of profitability based on the idea that AI will totally be used everywhere and by everyone, very soon now, you'll see. What finally prompted (pun intended) me to post this is the CEO yesterday unironically referencing AI 2027's "predictions".

Can't wait for the bubble to burst. I'm really curious to see if I'll keep my job through that. At the end of the day, the stuff I work on luckily has nothing to do with AI, and basically every other application of the product makes more sense; but now the entire company has shifted gears towards AI...

Cloudflare sponsors Ladybird and Omarchy, techfash workfare

They banned the guy that wrote the theil antichrist notes.

https://sfstandard.com/2025/09/23/spilled-peter-thiel-s-antichrist-secrets-now-s-banned-lectures/

Stephens and Thiel did not respond to requests for comment. Kulkarni declined to answer questions about the lectures, citing the off-the-record policy.

legal threats?

it would be very funny if Rocco Basilico’s legacy was that his name bore resemblance to Roko Basilisk and nothing else

Just got back from the Ted Chiang talk at the law school, talk was good but all the Q&A was lawyers ask-telling about LLMs. Not a single question for him about his fiction. :(

Bethany Brookshire:

Some people think ChatGPT has a place writing things like news briefs, stuff written to a specific style and tone and, y'know, kinda boring.

So Science did a study.

ChatGPT failed.

Why? It got stuff wrong. "Also, extensive editing for hyperbole was needed." https://www.science.org/content/blog-post/can-chatgpt-help-science-writers

The Nerd Reich has a post with criticisms of Thiel's Antichrist lectures, including some from European Catholic theologians.

Starting this Stubsack off with Brian Merchant's newest article: The Luddite Renaissance is in full swing

Cloudflare seems to be trying to make a shitcoin https://netdollar.cloudflare.com/

From RationalWiki: Yud claims that the only women he gave orgasms for completing math homework was his future wife. If he ever denied dating / playing with people from his foundation or making people who wanted to play with him fill out an IQ test I can't find it.