fjaraya retweeted
hey sam, your mclaren is only used 5% of the time. this is a terrible allocation of scarce resources. have you considered taking the bus to work
sam: local inference is a terrible allocation of scarce resources
“a lot of people want local inference, and somehow they assume that the problem is solved now because we have open-source models, we're going to be able to run them on our computers, and so we're not going to need all the stuff in the data centre
but from a global allocation perspective, actually it's quite terrible to have everybody, like, have a bunch of the world's limited supply of DRAM sitting in their own computer and being used five percent of the time
you want maximum utilisation in the cloud if you're trying to deal with this very supply-constrained environment
somehow many people's intuition is the opposite of that”
Hola @totalplaymx se nos fue el internet hace como dos horas, ya levante reporte, como le doy seguimiento?
fjaraya retweeted
Such a vibe shift happening right now. We've gotten nonstop requests from people who want to start using GLM-5.2. Incredible few weeks for open source models.
fjaraya retweeted
How to keep AI spend flat while token usage grows exponentially: Not with friction and spend alerts. With better defaults, routing, and caching.
Better Defaults (not Usage Caps) – Engineers can choose any model they want, but defaults matter. We’re experimenting with defaulting to open weight models like GLM 5.2 and Kimi 2.7 through our LLM gateway, while still encouraging engineers to choose the right model for the task. 91% of our employees were never hitting their usage caps, so instead of lowering caps and driving up alerts, we're moving to cheaper defaults. Note that code reviews use a diversity of models, so they can check each other's work.
Better Routing – In our custom harnesses, we preprocess prompts and route to the best model for the job, considering cache hits and model pricing. For instance, you may want a frontier model for planning, but not for execution where they can be overkill. Ultimately, humans shouldn't be choosing models - AI can automate this task.
Better Caching – Cache misses are the easiest way to drive your cost up. All of our requests are cache aware, so we’re reusing a warm cache wherever possible. For example, our cache hit rate went from 5% → 60% in LibreChat once properly implemented.
Keep Context Lean – Start fresh sessions when switching tasks. Scope file context narrowly. Disconnect unused tools. Don't just compact. The goal isn't fewer tokens used, it's fewer tokens wasted.
Better Visibility – Our engineers can use as many tokens as they want, from whatever model they want, but we’ve made usage visible – and the more you spend on AI, the more impact we expect.
The goal isn't to suppress usage. It's to build the infrastructure that makes exponential growth sustainable.
Putting this into practice has cut our AI spend nearly in half, while our token usage continues to grow.
fjaraya retweeted
Un tipo grabó miles de conciertos, los digitalizó y los subió a archive.org para que cualquiera pueda escucharlos.
Para esto se inventó internet, para compartir y no para hacer videos abriendo paquetes de Temu, pelotudxs de mierda.
archive.org/details/aadamjac…
From early Nirvana to Phish, a Chicago fan’s secret recordings of 10,000 shows are now online. blockclubchi.co/4t6zW7R
fjaraya retweeted
This innocent #cow was blindfolded to lull him into a false sense of security before cutting his throat. His friends are terrified and tried to offer comfort.
I'll never again pay others to do this stuff to animals. 😢 [photo: @iman.hamikhah / @animalsavemvmt]
I wear this crown of thorns
Upon my liar's chair
Full of broken thoughts
I cannot repair
If I could start again
A million miles away
I would keep myself
I would find a way
1
53
fjaraya retweeted
Science doesn’t need to go according to plan; it just needs to lead to a discovery. If doesn’t have to be done alone or together with a buddy; there just needs to be a discovery. It doesn’t need to happen fast or slow; just as long as there’s a discovery, then everybody is happy.
Inside the funhouse mirror factory: How social media distorts perceptions of norms sciencedirect.com/science/ar…
34
The Report of My Death Was an Exaggeration - freebsdfoundation.org/blog/t…
24
fjaraya retweeted
VERY unpopular opinion humans were never meant to be this busy that’s why everybody is anxious and depressed.
fjaraya retweeted
“Son animales sintientes”: Colegio Médico Veterinario de Chile critica venta de langostas vivas en supermercado
cnnchile.com/pais/son-animal…
fjaraya retweeted
"Puertas Abiertas @ESO_Chile -ALMA Santiago"
🗓️Viernes 21 de marzo
⏰16:00 a 20:00 hrs
¡A T E N C I Ó N!
🎟️Se abrirán los cupos los días:
Lunes, 17 de marzo, 16:00
Martes, 18 de marzo, 11:00
Miércoles, 19 de marzo, 10:00
Jueves, 20 de marzo, 13:00
🔗almaobservatory.org/es/anunc…
fjaraya retweeted
se cortó la luz. Algunos rápido quisieron culpar a las renovables. Para mí fueron una solución. Al cortarse el sistema solar con baterías se encendió automáticamente. En ese momento tenía una batería de 10kwh. Eso permite un consumo normal de la casa (tele, computadores, iluminación , refrigerador) funcionar por 20 horas. Tengo autogeneracion solar desde el 2012. Desde entonces los costos han bajado en 75% siendo Chile uno de los lugares donde más barato te instalan estos sistemas . En lugares con subsidios termina costando incluso más con subsidios.
fjaraya retweeted
En Football Manager no hacemos trampa ☺️
Caso Sartor: CMF oficia a Michael Clark para entregar información sobre toma de control de Azul Azul df.cl/mercados/fondos-de-inv…





























