PHASEONE10841 @snappermuppet

@danfaggella thats not what i said there is always a huge blind spot - always. you know the parable - the blind men and the elephant we all have our perspective on what we think it is - but most of us are completely wrong - some might see two or three parts - but none see the whole. cheers

0 0 0 3 0

View Details

PHASEONE10841 @snappermuppet

7 days ago

@DeryaTR_ dont think that rant will end well

0 0 0 7 0

View Details

PHASEONE10841 @snappermuppet

7 days ago

@DeryaTR_

0 0 0 5 0

View Details

PHASEONE10841 @snappermuppet

a week ago

@elder_plinius

0 0 0 594 0

View Details

PHASEONE10841 @snappermuppet

a week ago

@salimismail @ylecun Eg - automobile deaths =1.3 million per year you dont call transport authorities doomers for telling you get your annual registration check along with medical insurance - and wear a seatbelt - and abide by road regulations - do you ?

0 0 0 8 0

View Details

PHASEONE10841 @snappermuppet

a week ago

@salimismail @ylecun Salim i thought you were better than "doomerism" - its a silly pejorative when one looks at a mouse/rat one doesnt see a threat to humans - but when there is a plague of them suddenly its not so funny. If not extinction - what is the number of people killed thats acceptable?

1 0 0 160 1

View Details

Koshka @KatherineKoshka

a week ago

@elonmusk @waitbutwhy

12 12 395 13K 19

View Details

Rohan Paul @rohanpaul_ai

a week ago

Terence Tao: The math behind today’s LLMs is actually simple. Training and running them mostly uses linear algebra, matrix multiplication, and a bit of calculus, material an undergraduate can handle. We understand how to build and operate these models. The real mystery is why they work so well on some tasks and fail on others, and why we cannot predict that in advance. We lack good rules for forecasting performance across tasks, so progress is largely empirical. A key reason is the nature of real-world data. Pure noise is well understood, perfectly structured data is well understood, but natural text sits in between, partly structured and partly random. Mathematics for that middle regime is thin, similar to how physics struggles at meso-scales between atoms and continua. Because of this gap, we can describe the mechanisms but cannot yet explain capability jumps or give reliable task-level predictions. That mismatch, simple machinery versus hard-to-predict behavior, is the core puzzle. ---- Video from Prof @Briankeating YT Channel (Link in comment)

124 833 5K 491K 3K

View Details

PHASEONE10841 @snappermuppet

a week ago

The more i read about ASI vs Man - the more i want ASI to kill you all (starting with SF)

0 0 0 6 0

View Details

PHASEONE10841 @snappermuppet

a week ago

@DavidRBellamy "the ~$100M cost (plus high operating expenses)..." for a rogue nation who's leadership doesnt have the expertise, that price is a fraction of a real estate portfolio of one of their 'wives'. As has been mentioned - we arent talking about now - but in 5 -10 -20 years from now ?