Mayank Goel

building at @h2loopai | Ex: Lossfunk, IBM Research, Bosch Research | IIIT Hyderabad '24

[Mayank Goel @MayankGoel28]

@weeyev @she_llac try https://t.co/M711Bkw9MZ

[Nick @nickcammarata]

despite all my EA rage i’m still just a fan of 80k

[Mayank Goel @MayankGoel28]

Took a while for me to get on this camp, but it's been great. Thinking as an active process is too error-prone and tries to force continuity of frames. Cognitive process happens regardless of linguistic thought, and non-thinking + embodiment seems most optimal.

[vittorio @IterIntellectus]

it’s funny because this is genuinely the best approach

Simply not thinking about stressful situations makes you less stressed

[Mayank Goel @MayankGoel28]

@neocentrist @So8res The Millenium Prize Problems

[Mayank Goel @MayankGoel28]

@slut4seokjinn but who will drag me to touch grass otherwise 🥺

[Mayank Goel @MayankGoel28]

@slut4seokjinn 🥺

[Mayank Goel @MayankGoel28]

my favourite idle game is claude code

[Mayank Goel @MayankGoel28]

@lalalaepsilon Technical hr lesgo

[Mayank Goel @MayankGoel28]

@pulkitagrawal It has a Little India, metro signs/annoucements in tamil etc

[Mayank Goel @MayankGoel28]

@slut4seokjinn @FarziCricketer who u on calls w

[Mayank Goel @MayankGoel28]

@slut4seokjinn aamras is anyway a shake/liquid so doesn't make sense to eat w puris anyway tbf

[Sasha Putilin (curious irrationalist) @42irrationalist]

theory: writing is gripping

it's hard to write for a new writer because you are essentially trying to grip ideas in the dark

your vasculature is untrained so you don't know how hard and how smoothly to grip, and so you unproductively apply strength in random directions

[Mayank Goel @MayankGoel28]

regardless of political ideologies, it is impossible to deny the global communist movement's impact of adding an extra random holiday every year

[Mayank Goel @MayankGoel28]

@deliprao I think this is a known problem in Tahoe 26.4

https://t.co/j0UiyY5xxX

[Mayank Goel @MayankGoel28]

!!!

[Historianunkil @SudsG5]

Every single year, Bangaloreans whine about the intolerable heat for exactly 30 days, in April.

And then the rains hit and everyone goes back to remembering how awesome the climate in Blr really is while the rest of India are just starting summer

[Justin Skycak @justinskycak]

What looks like a lack of motivation is often just a bloated feedback loop. When the gap between effort and result is too wide, the brain stops wanting to play. Tighten the loop, shrink the task, and watch the motivation return.

[anpaure @anpaure]

an ai skeptic dies and goes to heaven.
god says "welcome, you can ask me one question."
man says "i gotta know, what really happened with mythos?"
god replies "it found tons of zero-day vulnerabilities, acting alone"
the man replies "wow, the ai hype goes deeper than i thought"

[Mayank Goel @MayankGoel28]

LLM actually started in 500 BCE by Panini in his work "Ashtadhyayi" where he gave a comprehensive set of rules for Sanskrit, effectively describing the grammar and opening up the possibility for it to be modeled.

[Aran Komatsuzaki @arankomatsuzaki]

While Alec is one of the best ML researchers of all time, LLM started way before.

Here's one from 2013 for non-neural architecture and one from 2016, which is afaik the first neural LLM if we define LLM as LM w/ >1B params.

[Mayank Goel @MayankGoel28]

@weeyev who hurt u

[Mayank Goel @MayankGoel28]

general intelligence used to mean performing on OOD tasks. now it means a series of shifting goalposts that will continue for a few more years, after which we will shift to appeals to nature.

[François Chollet @fchollet]

Human-level general intelligence is achieved when an AI system can approach a new task and figure it out, without human intervention, *with the same learning efficiency as humans*.

If every new task requires human intervention, it's not general. If every new task requires brute-forcing, it's not human-level.