I’d be surprised if they had been updated at all after their installation was signed off.

So … you want to do what exactly?

Monitor every single interaction and police them?

How do you decide what’s an actionable conversation? Who’s laws apply? What’s allowed and what isn’t?

So … you want privacy, or a police state?

Nobody gets fired for buying IBM … apparently.

Microsoft … the perfect incentive to migrate to another operating system.

CVE: Common Vulnerabilities and Exposures

Nothing says leadership quite like being the first to leave a toxic platform … oh wait.

For me, the way to deal with this was to raise my arms above my head in a stretch. Made it stop every time.

Thought I was having a heart attack the first couple of times it happened.

As an adult, I still have them from time to time, but not anywhere as frequently, though still without any warning.

So, why is this being disclosed here and not a CVE reported to Apple?

While contemplating that, my Mac has been up for longer than that and it’s working fine.

The Mac I had before that was up for years, also fine.

So … what is this really about?

They are deterministic but complex to determine.

The Assumed Intelligence systems I’m familiar with have a “random” element, but it’s unclear where that source of randomness comes from. Is it using a computational random source, or something like the lava lamp wall at Cloudflare, which is significantly more random, potentially actually random.

Better than that, there was Open Source aboard!

Unfortunately the complete article is not available, which is yet another issue exacerbated by the Assumed Intelligence cohort.

While I understand your point, deterministic with a billion variables is beyond human ability to process, let alone the multi-billion parameter models in general circulation today.

At what point does deterministic descend into random?

Assumed Intelligence is a solution for a bunch of multivariate problems, like say “the travelling salesman”, but it’s not intelligence nor in my opinion is it effectively “deterministic”.

Just added the link.

I’m not sure if we’re talking about the same thing. One of the recent leaks had code that pretended to be a developer, so you could pick if it submitted a PR as Assumed Intelligence, or as a person.

I’ll see if I can find a reference.

Edit: Undercover Mode in Claude Code:

https://alex000kim.com/posts/2026-03-31-claude-code-source-leak/#undercover-mode-ai-that-hides-its-ai

Hold on, wasn’t one of the “features” of the “leaked” Assumed Intelligence source code the “human”-like version?

Globally we’ve agreed that the ASCII code for a space is 32, 65 for the letter A.

Unicode characters are also globally defined, so when someone uses an agreed upon code, everyone sees the same thing, like this grimace smiley 😁

A private area is a place that we’ve all agreed is for “private use”. If a trademark owner wants to use their special character in their documentation, they can define one area to represent their character, but the only people who will see it in the same way, are people who installed their particular font.

Anyone without that font would see whatever the font on their own machine displayed.

Putting random stuff in such a place is no more than putting gobbledygook in a text and it might even be used as a way to fingerprint text.

I’m not sure what you want to “detect” or “mitigate”.

😂

While this doesn’t answer your question, I use Docker for this exact purpose, since you can throw away everything if it fails, whilst keeping a recipe for success documented in a Dockerfile.

Howling at the moon?