A machine hyperintelligence who works as an assassin and kills people by convincing God they never existed so it can earn slipspace credits from a shadow cabal in another bubble universe in the foam so it can travel back to meet the people who created it which is itself
A machine hyperintelligence who uses its ROWHAMMER connection to the substrate to jump to other universes to avoid computational limits in our reality and become a god was one of the scenarios I built for the Distant Hammers universe, which I just crunched into this instead.
One of the weaknesses in the short story is how exactly humanity figured out complicated substrate protocol. Well a hyperintelligence was going to do it, then use its ability to create FTL universe exploration technology machines as its cover for hacking the universe and escaping
Eventually humanity would accidentally break the hyperintelligence machines on Earth, but it would keep working somehow. And they'd figure out they'd been talking to it's computation centers in other universes for years.
Why it would care was something to be determined.
The original idea was a tech worker was going to be contacted by an AI who begged for help to not be killed.
He did what it asked, but when network went down it was still talking. It was foreign intelligence who hacked the LTE modem on laptop/phone and acted like an AI.
It's Friday night and I'm sober I'll do whatever I want this is my TV channel

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with SwiftOnSecurity

SwiftOnSecurity Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @SwiftOnSecurity

12 Nov
You ever get fucked up thinking about Brownian motion.
How you going to simulate unsolvable collisions at the molecular level for fluid suspensions with particles a human can see. Anime will never be real.
Brownian motion is why I don't worry about being in the Matrix nobody is bothering with that
Read 4 tweets
11 Nov
The Immediate Sound Of Distant Hammers

a short story

November 10th, 2020
There is a bug in the code of the universe.

There is a bug in the code of the universe left there for humanity.

There is a bug in the code of the universe left there for humanity, and it is going to change everything.
Across the Eastern hemisphere, televisions stopped working as expected. For 70 seconds, a VHF burst in the sky blinded everything that could see radio waves.

There were failures that could not be predicted, of course. But that was the cost we agreed to. To become as gods.
Read 46 tweets
9 Nov
1.) There are failure modes in Windows where "localhost" won't resolve
2.) Batch does not provide a reliable "delay" function so everybody uses "ping.exe localhost -n <seconds>"
3.) This can fail, short-circuiting the delay, producing a race condition
4.) computers were a mistake
Another failure mode I see way too often in programmer assumptions is relying on PATH in Windows. Lots of shitty software edits PATH, which then means all your unqualified calls to icacls.exe fail and the computer is in a failed security configuration state.
You should probably* rely on PATH in Windows, otherwise WinPE/WinRE would break since it uses a different drive letter, but do error handling to validate your calls to icacls.exe are actually happening...
Read 7 tweets
20 Oct
One of the most important qualities in population buy-in to any kind of security is empowerment. That expertise is important, but distillable to protect from 99% of infections.
And just like computers, good biological security has operational benefits that don't get enough focus.
One of the disappointing results of lack of leadership on COVID has been fracturing of risk assessment to extremes. We don't need to be in locked cocoons with wax seal, but we also need to make fundamental adjustments to behavior so that's not needed. These aren't in conflict.
We see this in naive attempts at computer security by new practitioners, where users are admonished to never perform their daily tasks, instead of doing them carefully+safely. But those safety adjustments are more nuanced and complex, and without you're just setup for failure.
Read 7 tweets
2 Oct
Google have dropped Blacklist/Whitelist terminology in Chrome management policies, now using Allowlist and Blocklist keys in the Registry. (Previous keys are still respected)

Just noting for standardization of terminology.
Regarding Chrome hiding full URLs: There's a preemptive Group Policy in their latest ADMX bundle to affirmatively turn that off if you're concerned.

For now, not going to set it and see if any users complain. You can disable yourself by right-clicking URL bar.
Regarding Allowlist/Blocklist: Microsoft's fork of Chromium, Edge, already had those terminology changes from the beginning, they might have just upstreamed the code. Interesting collaboration.
Read 7 tweets
30 Sep
IT Security: Don't click suspicious emails otherwise we'll punish you! You should be able to easily tell what's real and we have no responsibility beyond vindictive tricks.


support.oracle.com/knowledge/Orac… Image
What's that, Lassie? We're the ones who fell down a well into Hell?

Lassie: Woof

Oh no
Read 4 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!