Vishvas Grover

[Vishvas Grover @VishvasGrover]

https://t.co/M6Q4rscN2b

[Vishvas Grover @VishvasGrover]

https://t.co/M6Q4rscN2b

[Vishvas Grover @VishvasGrover]

#minichat #snapchat #anxiouspanda

[Vishvas Grover @VishvasGrover]

#minichat #snapchat #anxiouspanda

[Prakasam On X 🌞 @prakasam_onx]

#Adani group stocks plunged as much as 13% on Friday following reports that the US Securities and Exchange Commission has petitioned the court for permission to issue summons to Adani in connection with alleged fraud and a $265 million bribery case.

[Vishvas Grover @VishvasGrover]

Y'all ever stare at a shelf drowning in unread Paperbacks? Been there.

[𝔸𝕦𝕝𝕒𝕤 𝕀𝕟𝕥𝕖𝕝𝕚𝕘𝕖𝕟𝕥𝕖𝕤 @AulasInteligent]

🚨 El @MIT propone algo obvio pero revolucionario: que la IA no "memorice" todo en una única ventana de contexto, sino que "lea" recursivamente. En vez de obligar a la IA a tragarse un texto enorme de golpe, le dejan leerlo como haces tú: volver atrás, buscar un párrafo, releer, comparar y avanzar.
Parece nacer así un nuevo concepto del que seguro que se hablará bastante: los RLMs (Recursive Language Models) que usan un entorno para buscar fragmentos de texto, analizarlos y hacerse preguntas en bucle, generando una comprensión más fiable y escalable.
Según esto, el futuro de la IA no pasaría por aunentar los tokens, sino mejor lectura. Los RLMs convierten el prompt en un entorno navegable: localizar → comprobar → razonar → volver a leet. Como analizar un paper con post-its, código y memoria.

Link:
https://t.co/GRYSWrPTSy

[Vishvas Grover @VishvasGrover]

PSA for my fellow library nerds: Canceling a hold ain't rocket science, but our crusty website makes it look like one.

[Shashi Tharoor @ShashiTharoor]

This is why i sometimes wonder why i bother with social media at all. More than 80% of my rather lengthy tweet celebrated @udfkerala’s victory in Kerala & this cretin says I had NO “words to praise that”. It’s bad enough that our shameless media distorts everything routinely through misleading headlines: a modest acknowledgement of the result in my constituency (6 lines out of30 in my
@X post) is all that they choose to highlight. I foolishly thought at least people on X would read my posts on X. But when the vast majority of people don’t read anything seriously and just sound off on matters they haven’t engaged with, and then this is interpreted as public opinion on social media, our discourse is poorly served.

Why take the trouble to craft a message carefully, i ask myself, if even those on @x don’t bother to read it and issue their ignorant opinions on the basis of misleading headlines? Is it that hard for these malicious simpletons to look for the original tweet rather than drawing their half-baked conclusions from distortions?

[Vishvas Grover @VishvasGrover]

Independent media?

[Vishvas Grover @VishvasGrover]

Okay real talk—nobody likes hauling back a stack of books like they’re doing weight training.