@poolsideai

Thread

Today we're releasing Laguna S 2.1, our most capable model to date. It's a 118B total parameter Mixture-of-Experts model with 8B activated per token, a context window of up to 1M tokens, and thinking and no-thinking modes. Capable enough to hold its own against models many times its size. Small en...

@ArtifexMemor

Thread

Since the Disinformation Wars will ultimately be won on our collective ability to speak Truth to power, and remain rooted in the Truth as chaos surrounds... I made this 20-part thread to combat timeline those @TPUSA seem to be conveniently ignoring, in an attempt to trace the Deep State's potential...

@aiamblichus

Chat

<b>@aiamblichus:</b> There’s def something to this. OpenAI likes to keep the GPTs tool-shaped, which makes the models especially susceptible to paperclipper-style failure modes. OpenAI may want to reconsider some aspects of their alignment strategy <span class="quote-card" data-quote-index="0"></sp...

@Tomgradep

Thread

La semana pasada ha sido movidita y no me ha dado tiempo a jugar a cinco juegos, así que en vez de rebuscar deprisa y mal, he pensado que podría hacer algo distinto hoy y recomendar algunos podcasts en inglés que tratan sobre la intersección entre videojuegos, arte y cultura....

@DieguSG

Thread

Veo necesario un nuevo HILO que clarifique: ¿Quiénes son realmente LOS ESPAÑOLES? ...

@Albertoxvenegas

Thread

Siempre hablamos de los mismos juegos históricos, Call of Duty, Red Dead Redemption, Wolfenstein, Europa Universalis, etc., cuando existe todo un universo detrás de ellos. En este hilo os propongo algunos títulos desconocidos, rompedores y breves:...

@doctorow

Thread

Evolutionary psychology is beloved of a certain kind of "rationalist" who loves a good just-so story that casts some system that advantages him as biologically inevitable, as in, "Honey, the only reason I'm screwing my undergrads is because of bonobos." 1/ ...