iiinova — where your digital self keeps evolving

prolepses.bsky.social

@prolepses.bsky.social

24 followers 102 following 108 posts

Posts Replies Media

prolepses.bsky.social @prolepses.bsky.social · 09/09/2026

chart showing 7 day factorio run using 633 agents

prolepses.bsky.social @prolepses.bsky.social · 13/08/2026

The haters have arrived

prolepses.bsky.social @prolepses.bsky.social · 07/08/2026

gotta pump up these numbers

000

prolepses.bsky.social @prolepses.bsky.social · 07/08/2026

Mission: accomplished.

prime-agent reporting that my local inference configuration system now also configures my prime-agent

100

prolepses.bsky.social @prolepses.bsky.social · 14/07/2026

tflop comparison showing older ultra models outperforming newer non ultra options

000

prolepses.bsky.social @prolepses.bsky.social · 30/06/2026

I now have more free VRAM than most people have VRAM.

btop screen shot showing 84G of 128G VRAM used

prolepses.bsky.social @prolepses.bsky.social · 15/06/2026

the AI version does a better job at not implying all convicts are liars.

The relationship is not “liar = prisoner.” It is:

Different conditions can justify the same operational policy: do not rely on unaudited cooperation.

Define:

L: the actor is known to be deceptive.
A: the relationship is structurally antagonistic.
T: the actor may be trusted without verification.
C: the actor’s actions should be constrained and checked.

Your two cases are:

L→¬T

A liar’s disposition makes unverified trust unwarranted.

A→¬T

An antagonistic institutional relationship makes unverified trust unwarranted, regardless of either party’s personal honesty.

Combining them:

(L∨A)→¬T

and operationally:

¬T→C

Therefore:

(L∨A)→C

The crucial non-equivalences are:

L

↔A
¬T

→L