so_tired

See my other comment

> Humans get a hand crafted curriculum inputs, evolved over 1000s of iterations, in a near-optimal language encoding.

> Human reading/talking/listening equivalent of 200 pages of text per day for 80 years would be just 13GB of raw data or 3B tokens

I am sympathetic to the "total life time input" argument.

But humans get a hand crafted curriculum inputs, evolved over 1000s of iterations, in a near-optimal language encoding.

Also, if unsupervised gets us in-dist, and DRL seems to be not-bad in search-out-of-dist.... then we are getting close ?

Certainly a x100 scaling of current techniques can get a useful enough machine that makes many human tasks trac-able?

(I am not getting into the Turing / AGI / skynet argument)

The title and quote are obviously hyperbolic.

But

A large portion of that mass is in a too high or too low temperature state to support any complexity (I am not a physicist!)

On the other hand, a single pound of neurons is more powerful than a warehouse of nano-scale-minituarized silicon.

Consciousness itself may or may not be a purely emergent property of this complexity.

We already have fairly robust (and simple!) rules for most physical, chemical and biological behaviour.

Other than the creation and ending of he universe, Consciousness is certainly a top candidate no ?

I appreciate Billionaires that take city-size estates out in the country where there is plenty of space.

I feel this is a lot less controversial than urban mega-dwellings which can cause undue hardship for mere millionaires.

9front looks so very interesting. Thanks

Ugh! Thanks but no thanks

I dont want to mount and expose nfs, or allow a full ssh login. This will be a custom agent/daemon on both ends

Which lib/approach would you recommend if i need run a remote batch process using local files?

For example, on the local system:

remote-execute gcc localfile.c -o localfile.o

And on the remote system, there is a gcc and a remote agent which receives and retransmits the files

Do they need to optimize for reduced memory/communication patterns? Or is that going too far

AFAIK, the memory-communication power can increase by orders of magnitude if u go off-core, off-chip, off-cache, etc.

Do they need to optimize for reduced memory/communication patterns? Or is that going too far

AFAIK, the memory-communication power can increase by orders of magnitude if u go off-core, off-chip, off-cache, etc.

Why is this getting flagged? Are large tech companies now somehow shielded from criticism?

This experience actually sounds both common and unavoidable.

FAANGs have x10 candidates for every position. So the interviews are indeed both difficult and arbitrary. And after a rejection they can always call you after 6 months and restart the whole shtick. So why not. Sigh.

As someone who interviewed 100s of software engineers - its just the system doing its best, and there is usually no bad-intent on either side.

Important line from the article

> a practice at odds with the company’s stated policies...

> .. as stated to congress

Can you give some salaries for hiring tech workers?

2-3 years experience, web front end and/or B2C marketing

Is there any research that spiking/dopamine type learning is good at "animal level" behaviour, but abstract and complex thinking is enabled by different mechanisms ?