orena

But what to do with android YouTube app ??

I do the following: www.yandex.com MOVIENAME direct streaming

And that is it... I have prime + netflix + hulu and such, but I use "yandex.com" as it does not have ads - even if sometimes it takes a bit to load and rarely it gets stuck for a second, it is less time than the stupid ads.

Any results on frontier math or arc ?

How many epochs did you train with ? 100k hours is not a lot for an LLM, Feels like bitter lesson

Although very cool, I lack the emotional context to understand why ppl work on it, what is the motivation that drives them ?

Not trying to offend, just trying to understand

Q (Cue) | Boston, MA | Onsite | Full-time | AI/ML – Speech & Audio

We’re building a revolutionary new way for people to communicate.

Our work is deeply rooted in large-scale deep learning. We train massive models across many GPUs using proprietary data, pushing the boundaries of what we can do with deep learning.

We’re looking for someone who is both highly research-oriented and technically strong. Training large models over long durations demands not only cutting-edge research thinking but also solid engineering skills and deep experience in scaling DL systems.

https://q.ai/open-position/?gh_jid=4569618101

I've asked claude to explain what you meant... https://claude.ai/share/391160c5-d74d-47e9-a963-0c19a9c7489a

So no bitter lesson?

The only reason IL can do it, is because the general population in Lebanon hates hezbollah, it is basically a proxy control power from IRAN.

Look at this to understand how big part of Lebanon sees Hezbollah https://www.instagram.com/brigitte_gabriel/reel/CyT_RanK1gi/

Page. 57 in Real Analysis, definition of a limit...

First time in my life that I did not understand something on the spot (actually took me a few months to (REALLY) understand it).

The style is not very different than NeMo(nvidia)/fairseq(Facebook)/espent(oss) etc..

They publish the source code for Whisper...

Come on, no one knows... after 10 years in computation neuroscience + experimental neuroscience. The slope of neuroscience advances in very very low (some will say negative). The slope of AI advances in much much higher. --> we will get an AI to understand the brain an explain to us, it will not come from a lab.

Just my point of view

Very cool! Now Conv3D..

I always wonder how to present a realistic Conv3D, as with channels it is actually 4D, it is an interesting challenge.

qdirstat - https://github.com/shundhammer/qdirstat

Have you seen orcam projects ?

Probably yes, it works on 2p Neuro images, which are of course not InFocus

Turboreg is also good: Paper: http://bigwww.epfl.ch/publications/thevenaz9801.html

Python implementation: https://pypi.org/project/pystackreg/

Is there a podcast of their talk?

Well, you don't know how many years humans will need in order to solve it ;-).

It might be that the Riemann Hypothesis will be the first of Math's "Hard Problems" that will be solved by AI before humans...

For those that think that Bitcoin is not environmental, please read -https://bitcoinmagazine.com/business/open-letter-bitcoin-tes...

There is no logical fallacy, I did not said that it will continue to go up more that it goes down, I just said that history shows...

Sometimes it goes down, sometimes it goes up - it's not the first time. History shows that it goes up more than it goes down.

There is a solution to that, if they remove the making on channels in/out funds the problem is solved...

Not true - https://github.com/breez/LightningRod

Amazing software, great development team. First I replaced Dropbox with resilio and a year ago replaced resilio with synchthing