AI Models​ 10.000 Times the size of GPT-4 are possible by 2030

[John Clark]

And during this key moment, the most critical in the entire history of Homo sapiens, Donald Trump will be the most powerful member of that species.

AI Models Scaled Up 10,000x Are Possible by 2030, Report Says

John K Clark    See what's on my new list at  Extropolis

mtk

[Cosmin Visan]

So what ? I haven't seen a single useful use of AI. Can you name 1 ?

[John Clark]

On Thu, Dec 19, 2024 at 2:29 PM 'Cosmin Visan' via Everything List <everyth...@googlegroups.com> wrote:

> So what ? I haven't seen a single useful use of AI. Can you name 1 ?

I can name a few more than one. AI has found the structures of 200 million proteins.

John K Clark    See what's on my new list at  Extropolis

j34

 

[Cosmin Visan]

Yeah, sure. =))))))))))))))) Post the link.

[John Clark]

On Thu, Dec 19, 2024 at 3:16 PM 'Cosmin Visan' via Everything List <everyth...@googlegroups.com> wrote:

>>> So what ? I haven't seen a single useful use of AI. Can you name 1 ?

>> I can name a few more than one. AI has found the structures of 200 million proteins.

> Yeah, sure. =))))))))))))))) Post the link.

The latest database release contains over 200 million entries 

 

John K Clark    See what's on my new list at  Extropolis

3c1

[Cosmin Visan]

Link please!

[John Clark]

On Fri, Dec 20, 2024 at 3:32 AM 'Cosmin Visan' via Everything List <everyth...@googlegroups.com> wrote:

Link please!

As you can see I already did that yesterday! 

[John Clark]

[Cosmin Visan]

Sorry, I didn't know that blue text is actually a link.

[Cosmin Visan]

Ok, I entered the link. The main title is:

"AlphaFold DB provides open access to over 200 million protein structure predictions to accelerate scientific research."

You probably don't understand what this title means. This title means that AI made 200 million random predictions and now it's up to the researchers to verify them and see if they are good. So it didn't help at all. I can equally make random predictions and ask others to do the real job of verifying.

[John Clark]

On Fri, Dec 20, 2024 at 6:01 AM 'Cosmin Visan' via Everything List <everyth...@googlegroups.com> wrote:

> You probably don't understand what this title means. This title means that AI made 200 million random predictions and now it's up to the researchers to verify them and see if they are good. So it didn't help at all. I can equally make random predictions and ask others to do the real job of verifying.

The people who made the AI that produced those 200 million structures for proteins just received a Nobel Prize for doing so!! But I suppose you'll just say the Nobel Prize does not exist. 

Cosmin, I always knew you were an ignoramus but until this moment I didn't know that you're so stupid you don't realize how world-class stupid you are. I have no problem with stupid people, but I do have a problem with people who are so stupid they don't know they're stupid, because then they don't even try to learn.

John K Clark    See what's on my new list at  Extropolis

0jx

[Cosmin Visan]

There are 1000 biologists in the world. This would mean 200.000 verifications per biologist. If he makes 1 verification/day, it would take them 547 years to verify those predictions. At a rate of 1 verification/hour in a 8 hours work-day, it would take them 68 years to verify them. So what are you even taking about ? Hallucinating much ?

Also, link to that Nobel prize please, so I can teach you how to properly understand the hyped news, because you clearly lack the skill of accurately reading words.

[John Clark]

On Fri, Dec 20, 2024 at 8:27 AM 'Cosmin Visan' via Everything List <everyth...@googlegroups.com> wrote:

> Also, link to that Nobel prize please

For Darwin's sake! Do you know how to use Google? 

John K Clark    See what's on my new list at  Extropolis

ggg

The people who made the AI that produced those 200 million structures for proteins just received a Nobel Prize for doing so!! But I suppose you'll just say the Nobel Prize does not exist. 

Cosmin, I always knew you were an ignoramus but until this moment I didn't know that you're so stupid you don't realize how world-class stupid you are. I have no problem with stupid people, but I do have a problem with people who are so stupid they don't know they're stupid, because then they don't even try to learn.

John K Clark    See what's on my new list at  Extropolis

0jx

--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/everything-list/db1f91a4-511c-47d9-b5c2-ed49abe52ea9n%40googlegroups.com.

[Quentin Anciaux]

John, it's useless... as much as dislike sometimes your way to argue, you're a real human being with whom discussion is possible, those guys aren't. It's useless, they have the world all sussed out, will only reply insults, will change the conversation when they want, their pleasure is to annoy people. Shame they found this list. I missed Wei dai, Schmidhuber, finney, Bruno marchal. 

To view this discussion visit https://groups.google.com/d/msgid/everything-list/CAJPayv1QMOMfD%2BPT1pFahDDq7sJzy1MmgLRHvc4w661XwumB_w%40mail.gmail.com.

[Cosmin Visan]

I know how to use google. The problem is that you don't know to read words. As such, in those links it says "1+1=2" and you read "231256*4234324 = 312312321! ZOMG!". So I have no idea what Nobel prize are you referring to, since clearly there was no Nobel prize for what you say. So please give me the link and I will translate to you exactly what it says there.

[John Clark]

On Fri, Dec 20, 2024 at 9:50 AM 'Cosmin Visan' via Everything List <everyth...@googlegroups.com> wrote:

> I have no idea what Nobel prize are you referring to, since clearly there was no Nobel prize for what you say.

 

OK, this is how you use Google. If you look down you will see a square thing, it's called a "keyboard". You can type on it. In your browser type in "Google.com". Once you're on Google's website type in "AlphaFold Nobel Prize" and the very first thing that will pop up is the following, and yes, the blue color means it's a link:

 

Chemistry Nobel goes to developers of AlphaFold AI that predicts protein structures

John K Clark    See what's on my new list at  Extropolis

3ew

[Russell Standish]

On Fri, Dec 20, 2024 at 03:32:24PM +0100, Quentin Anciaux wrote:
> Shame they found this list. I missed Wei dai, Schmidhuber, finney, Bruno
> marchal. 

Speaking of Schmidhuber, he's been busy lately dissing the Nobel prize
given to Hopfield and Hinton for
ANNs. https://people.idsia.ch/~juergen/physics-nobel-2024-plagiarism.html

And he has a point - given he discovered the LTSM RNN algorithm, he'd
know a thing or two about the history of ANNs.

Cheers


--

----------------------------------------------------------------------------
Dr Russell Standish Phone 0425 253119 (mobile)
Principal, High Performance Coders hpc...@hpcoders.com.au
http://www.hpcoders.com.au
----------------------------------------------------------------------------

[Cosmin Visan]

They are just retards. What's the big deal ? See here:

https://www.reddit.com/r/labrats/comments/1b1l68p/people_are_overestimating_alphafold_and_its_a/

"And Alphafold doesn’t produce protein structures. It produces structure predictions."

And I told you. I can predict that Santa Claus lives at the north pole. Now YOU go and prove it! In the meantime, please give me the Nobel prize for doing absolutely nothing, and YOU go and do the real work!