/sci/ - Science & Math » Thread #11883580

657KiB, 817x761, Researchers build robot scientist that has already discovered a new catalyst.png

View Same Google iqdb SauceNAO

Anonymous Thu 09 Jul 12:10:26 2020 No.11883580 View Reply Original Report

Quoted By: >>11883614 >>11883654 >>11884125 >>11884198 >>11884283 >>11884441 >>11884529 >>11884668

>Lab rats on suicide watch

Anonymous

Anonymous Thu 09 Jul 2020 12:34:34 No.11883614 Report

Quoted By:

>>11883580
jokes on you
my lab can't afford a proper heating mantle

Anonymous

Anonymous Thu 09 Jul 2020 13:00:51 No.11883654 Report

Quoted By: >>11883663 >>11883684 >>11883696 >>11884125 >>11884415

>>11883580

It's going to be a full time job for someone to maintain that robot. I do think tasks like mixing random combination of chemicals should be automated tho.

Anonymous

Anonymous Thu 09 Jul 2020 13:04:51 No.11883663 Report

Quoted By:

>>11883654
>t's going to be a full time job for someone to maintain that robot.
It's going to be done by the rats.

Anonymous

Anonymous Thu 09 Jul 2020 13:17:10 No.11883684 Report

Quoted By: >>11883983

>>11883654
While taking the jobs of many full time lab workers.

Anonymous

Anonymous Thu 09 Jul 2020 13:23:48 No.11883696 Report

Quoted By:

>>11883654
t. seething "scientist"

Anonymous

Anonymous Thu 09 Jul 2020 16:08:29 No.11883983 Report

Quoted By:

>>11883684
>While taking the jobs of many full time lab workers.

Perhaps. Im a robotics engineer and not a chemist tho.

Anonymous

Anonymous Thu 09 Jul 2020 17:01:47 No.11884113 Report

Quoted By:

>implying it can do meme shleck line shit

Anonymous

Anonymous Thu 09 Jul 2020 17:07:28 No.11884125 Report

Quoted By:

>>11883580
>already discovered a new catalyst
This is another one of those overhyped "AI can do smart things now" headlines that's basically trial and error, right?

>>11883654
>It's going to be a full time job for someone to maintain that robot.
This is a common pro-automation argument, but it's not like they're going to build enough robots that maintaining them is going to replace every job they take over in every context.

Anonymous

Anonymous Thu 09 Jul 2020 17:28:59 No.11884198 Report

Quoted By: >>11884478

>>11883580
so how did the robot even "discover" anything? We're just talking about a robot arm here, not AI right? In that case it's nothing new, there's lots of ways to automate experiments that are in use already.

Anonymous

Anonymous Thu 09 Jul 2020 18:07:38 No.11884283 Report

Quoted By: >>11884478

>>11883580
>built themselves
>KUKA
This is an off-the-shelf robot arm that does exactly what it's being told. Lame. But cool that lab rats are now pretty much obsolete and the good ones can be used for actual research.

Anonymous

Anonymous Thu 09 Jul 2020 18:52:32 No.11884415 Report

Quoted By:

>>11883654
>It's going to be a full time job for *someone* to maintain that robot.
plot twist: it's gonna be another robot

Anonymous

Anonymous Thu 09 Jul 2020 19:02:49 No.11884441 Report

Quoted By: >>11884448

>>11883580
Imagine you could build entire industrial material discovering full automated complexes. And there is no upper limit since there are so many random permutations you can assemble elements.

You just need to build also a machine who can harvest all the data. (cause i dont want anyone to have a job :D)

Anonymous

Anonymous Thu 09 Jul 2020 19:05:22 No.11884448 Report

Quoted By:

>>11884441
you know guys you pump the elements on one side in, it gets randomly assembled. assembly belt tested, and then you get an raw output what can do what.

Anonymous

Anonymous Thu 09 Jul 2020 19:15:50 No.11884478 Report

Quoted By: >>11884536 >>11884547 >>11884595

>>11884198
>>11884283
https://phys.org/news/2020-07-robot-scientist-catalyst.html

>The robot's brain uses a search algorithm to navigate a 10-dimensional space of more than 98 million candidate experiments, deciding the best experiment to do next based on the outcomes of the previous ones.
>By doing this, it autonomously discovered a catalyst that is six times more active, with no additional guidance from the research team.

Anonymous

View Same Google iqdb SauceNAO KE6QNYV.png, 1MiB, 1357x1281

Anonymous Thu 09 Jul 2020 19:40:01 No.11884529 Report

Quoted By:

>>11883580
learn2code, wetlab cucks.

Anonymous

Anonymous Thu 09 Jul 2020 19:43:57 No.11884536 Report

Quoted By: >>11884551 >>11884561 >>11884568 >>11884621

>>11884478
This is so fucking absurd

Can’t we simulate molecular chemistry enough to predict the results of such an experiment without a robot alchemist mixing fluids together in a box?

It seems that technology is inevitable and this is an absurd precursor wasting chemical reagents to maybe find results a few years sooner.

Anonymous

Anonymous Thu 09 Jul 2020 19:49:43 No.11884547 Report

Quoted By: >>11884595

>>11884478
Lmao, I remember reading the smug "researchers" here just few days ago badmouthing everyone and saying that "research will be the last thing automated if they can even do it".

Anonymous

Anonymous Thu 09 Jul 2020 19:51:18 No.11884551 Report

Quoted By:

>>11884536
A real man plow his field with his bare teeth. Its absurd how many people find so many straw-mans, to being anti progresses. Why you assholes dont fucking migrate to Afrika into the jungle, in your superior low tech world.

Anonymous

Anonymous Thu 09 Jul 2020 19:55:13 No.11884561 Report

Quoted By:

>>11884536
>"wasting" couple bucks worth of chemicals to get research done years sooner
I wonder what he means by this

Anonymous

Anonymous Thu 09 Jul 2020 19:58:04 No.11884568 Report

Quoted By: >>11884655

>>11884536
Seethe more
No, we cannot “simulate molecular chemistry” like that you autist

Anonymous

Anonymous Thu 09 Jul 2020 20:11:51 No.11884595 Report

Quoted By: >>11884622

>>11884478
Huh, okay, that's actually more impressive than what OP pic implies.

>>11884547
This. The problem really isn't that low-level jobs will be automated quickly, it's actually the high-level jobs like science and engineering that we'll lose to "AI". Robots can't declog your toilet, but they can automatically do research.

Anonymous

Anonymous Thu 09 Jul 2020 20:23:03 No.11884621 Report

Quoted By:

>>11884536
>Can’t we simulate molecular chemistry enough to predict the results of such an experiment without a robot alchemist mixing fluids together in a box?
no
>wasting chemical reagents to maybe find results a few years sooner.
i bet you're the type to cry about wasting taxpayer dollars on scientific research too

Anonymous

Anonymous Thu 09 Jul 2020 20:23:04 No.11884622 Report

Quoted By: >>11884688 >>11884690

>>11884595
>Robots can't declog your toilet
You’re thinking in terms of 4g

In the 5g world everything is connected, the Amazon Plumberbot will have access to your smart home network and will be able to dispatch from the nearest fulfillment center directly to your door let itself in find the most direct route to your bathroom and plunge the toilet the moment a clog is detected.

I can absolutely assure you, the obstacles in getting in front of the toilet are the majority of the technical challenge in designing a robot that can declog it.

Anonymous

Anonymous Thu 09 Jul 2020 20:33:19 No.11884655 Report

Quoted By:

>>11884568
*yet

Anonymous

Anonymous Thu 09 Jul 2020 20:39:02 No.11884668 Report

Quoted By:

>>11883580
This is a good thing though. Now we can do more science, faster. All researchers need to become technical anyways, and learn how to automate their work. For fields that are still rather undeveloped like tissue engineering, I don't see automation taking over research any time soon.
I am sick of reading shitty code from fellow researchers anyways.

Anonymous

Anonymous Thu 09 Jul 2020 20:43:15 No.11884688 Report

Quoted By:

>>11884622
Actually, a Plumberbot that roams through the sewers and pipes network doesn't seem a bad idea.

Anonymous

Anonymous Thu 09 Jul 2020 20:43:35 No.11884690 Report

Quoted By:

>>11884622
>I can absolutely assure you, the obstacles in getting in front of the toilet are the majority of the technical challenge
Yup. These obstacles don't exist for theoretical research and are much more tractable for experimental research. My point stands. "AI" can already design molecules, materials, stable structures, software etc, but robots can't build any of it so far.

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches