@jacobmparis An LLM can be deterministic: same weights + same input + greedy decoding → same output.
Non-determinism comes from:
• probabilistic sampling (temperature, top-k/top-p)
• floating-point & GPU parallelism
• batch / kernel scheduling
It’s controlled randomness layered on top.
@johnrushx Because we mistook language generation for intelligence.
LLMs optimize expression, not understanding.
Productivity grows when tools restructure cognition, memory and decision-making, not when they just talk better.
I don't know what I did to deserve a red flag 🚩. My activity is limited to a few likes and retweets, very few posts. In any case, I feel hyper-controlled, misunderstood, and helpless when faced with unexplained decisions.⬇️
@PhysInHistory If reality is a simulation, visible matter can be made of objects properties, dark matter can be made of hidden code, for example relations between objects. Dark energy can be the energy spent by the simulation to maintain relations between objects.
🚨BREAKING: Elon Musk has been nominated for a Nobel Peace Prize for protecting Free Speech
THIS IS MASSIVE 🔥
Simple poll. Please be honest! As of today, how much do you still trust this man?
A. 100%
B. 75%
C. 50%
D. 25%
E. 0%
Come una Salis qualunque, i vescovi scomunicano la stretta sui clandestini decisa dalla UE.
In Italia, oltre alle coop rosse dell’accoglienza, esistono pure quelle bianche.
@franborgonovo
Capito il perché di tutte quelle bandiere europee? Ai dipendenti dell’Unione europea che già prendono stipendi esentasse, gli hanno aumentato lo stipendio 8 volte in 3 anni. Nel frattempo in Italia gli stipendi sono fermi a 30 anni fa. Comoda la vita così eh?
Ennesimo boomerang finito sui denti del Pd che aveva subito grufolato all’attacco fascista.
“Avevamo bevuto troppo e abbiamo causato questi danni, non siamo in alcun modo coinvolti in organizzazioni filofasciste e non la pensiamo in quel modo”
Quindi, nessun raid fascista contro la sede del Pd a Chiavari.