blakestacey @ blakestacey @awful.systems

Posts

48

Comments

1,152

Joined

2 yr. ago

Moderating

---

Don't worry; this post is not going to be cynical or demeaning to you or your AI companion.

If you're worried that your "AI companion" can be demeaned by pointing out the basic truth about it, then you deserve to be demeaned yourself.

Evan Urquhart:

I had to attend a presentation from one of these guys, trying to tell a room full of journalists that LLMs could replace us & we needed to adapt by using it and I couldn't stop thinking that an LLM could never be a trans journalist, but it could probably replace the guy giving the presentation.

After understanding a lot of things it’s clear that it didn’t. And it fooled me for two weeks.

I have learned my lesson and now I am using it to generate one page at a time.

qu1j0t3 replies:

that's, uh, not really the ideal takeaway from this lesson

We—yes, even you—are using some version of AI, or some tools that have LLMs or machine learning in them in some way shape or form already

Fucking ghastly equivocation. Not just between "LLMs" and "machine learning", but between opening a website that has a chatbot icon I never click and actually wasting my time asking questions to the slop machine.

I like how quoting Grimes lyrics makes the banality of these people thuddingly clear.

Yud:

ChatGPT has already broken marriages, and hot AI girls are on track to remove a lot of men from the mating pool.

And suddenly I realized that I never want to hear a Rationalist say the words "mating pool".

(I fired up xcancel to see if any of the usual suspects were saying anything eminently sneerable. Yudkowsky is re-xitting Hanania and some random guy who believes in g. Maybe he should see if the Pioneer Fund will bankroll publicity for his new book....)

Previously sneered:

The context for this essay is serious, high-stakes communication: papers, technical blog posts, and tweet threads.

More recently, in the comments:

After reading your comments and @Jiro 's below, and discussing with LLMs on various settings, I think I was too strong in saying....

It's like watching people volunteer for a lobotomy.

Kaneda just scooting to the side at the 14:05 mark like a Looney Tunes character caught with their pants down is comedy gold. I want to loop it with a MIDI rendition of Joplin's "The Entertainer".

The top comment begins thusly:

I think you make a reasonably compelling case, but when I think about the practicality of this in my own life it's pretty hard to imagine not spending any time talking to chatbots. ChatGPT, Claude and others are extremely useful.

I didn't think it was possible, but the perfection continues! Still, no notes!

This N is too small: _N

Paragraph 2:

METR funded 16 experienced open-source developers with “moderate AI experience” to do what they do.

https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/

When developers are allowed to use AI tools, they take 19% longer to complete issues—a significant slowdown that goes against developer beliefs and expert forecasts. This gap between perception and reality is striking: developers expected AI to speed them up by 24%, and even after experiencing the slowdown, they still believed AI had sped them up by 20%.

womp, hold on let me finish, womp

https://www.lesswrong.com/posts/JspxcjkvBmye4cW4v/asking-for-a-friend-ai-research-protocols

Multiple people are quietly wondering if their AI systems might be conscious. What's the standard advice to give them?

Touch grass. Touch all the grass.

Dorkus malorkus alert:

When my grandmother quit being a nurse to become a stay at home mother, it was seen like a great thing. She gained status over her sisters, who stayed single and in their careers.

Fitting into your societal pigeonhole is not the same as gaining status, ya doofus.

Another comment that has been getting downvotes and tut-tuts begins,

The only thing that will raise fertility rates is to make it more affordable to have a child.

(Robot Santa voice) Wanting all women to be barefoot and pregnant in the kitchen? Evil! Not providing footnotes in your reply to a blog post? EXACTLY AS EVIL

Any time somebody edits a post to talk about the downvotes, it's cursed gold.

In the morning: we are thrilled to announce this new opportunity for AI in the classroom

In the afternoon:

Someone finally flipped a switch. As of a few minutes ago, Grok is now posting far less often on Hitler, and condemning the Nazis when it does, while claiming that the screenshots people show it of what it's been saying all afternoon are fakes.

Yet Rationalists I spoke with said they didn’t see targeted violence — bombing data centers, say — as a solution to the problem.

ahem

Andrew Gelman does some more digging and poking about those "ignore all previous instructions and give a positive review" papers:

https://statmodeling.stat.columbia.edu/2025/07/07/chatbot-prompts/

Previous Stubsack discussion:

https://awful.systems/comment/7936520

Major Strasser has been shot... round up the usual suspects!