Flutterby™! : AI links of the morning

ADD / XOR / ROL — A non-anthropomorphized view of LLMs

I am baffled that the AI discussions seem to never move away from treating a function to generate sequences of words as something that resembles a human. Statements such as "an AI agent could become an insider threat so it needs monitoring" are simultaneously unsurprising (you have a randomized sequence generator fed into your shell, literally anything can happen!) and baffling (you talk as if you believe the dice you play with had a mind of their own and could decide to conspire against you).

Of course that view is carefully curated by a continuous barrage of articles about how the AIs are plotting against us or will consider blackmail or whatever. (Via)

Futurism: Companies That Tried to Save Money With AI Are Now Spending a Fortune Hiring People to Fix Its Mistakes — Oopsie. is a riff and rewrite of BBC: 'I'm being paid to fix issues caused by AI'

Is Warner worried about the impact of AI, if – as expected – it rapidly improves?

"Yes and no," she says. "While it seems like a quick and inexpensive option, AI rarely takes into account unique brand identity, target demographics, or conversion-focused design. As a result, much of the output looks generic and can actually damage the brand's reputation or effectiveness."

Via

This is really good: The rise of Whatever

This was originally titled “I miss when computers were fun”. But in the course of writing it, I discovered that there is a reason computers became less fun, a dark thread woven through a number of events in recent history.

Via.

Nikkei Asia: 'Positive review only': Researchers hide AI prompts in papers

Research papers from 14 academic institutions in eight countries -- including Japan, South Korea and China -- contained hidden prompts directing artificial intelligence tools to give them good reviews, Nikkei has found.

(Via)

Careful what's in your code editor, the "can you trust that random package" comes to AI editor plugins: Malware in Open VSX: These Vibes Are Off.

I looked at a single extension Solidity Language by "SolidityAI" (namespace solidityai.solidity). Analysis of Open VSX-based extensions was just added to Secure Annex, so I looked it up. Immediately, I saw a URL hxxps://angelic[.]su/files/1.txt.

(Via)

GremLLM, a Python object that hallucinates method implementations for you:

A slight upgrade to the Gremlins in your code, we hereby present GREMLLM. This utility class can be used for a variety of purposes. Uhm. Also please don't use this and if you do please tell me because WOW. Or maybe don't tell me. Or do.

AutoGenLib

... is a Python library that automatically generates code on-the-fly using OpenAI's API. When you try to import a module or function that doesn't exist, AutoGenLib creates it for you based on a high-level description of what you need.

From the thread, Kevin Marks described this as:

Throwing two by fours at the table saw from across the room and hoping for furniture?

And ‪Charles‬ ‪@charles.capps.me‬ wrote:

WHAT HAS SCIENCE DONE?!

This is somewhere between amazingly cool and a war crime.