Jump to content
IGNORED

Google announces it's nearing quantum supremacy


Rubin Farr

Recommended Posts

I can't imagine how our world will change when quantum computing becomes the norm.  Machines that are millions or billions of times more intelligent than we are.  We will integrate them into our lives and depend on them like never before.  The rest of this century is gonna be an amazing time of change, for better and worse.

https://www.moneycontrol.com/news/world/google-announces-breakthrough-in-quantum-computing-research-4568371.html/amp

Link to comment
Share on other sites

  • Replies 57
  • Created
  • Last Reply
16 minutes ago, Rubin Farr said:

can't imagine how our world will change when quantum computing becomes the norm.

This video provides a clue: 

 

Link to comment
Share on other sites

interesting how they need to design a fault-tolerant computer around it. i want to learn more about this.

 

i listened to a podcast a while ago with a guy from DOD who was talking about how quantum computing could nullify encryption "cryptosystems" that military, finance, corporate, and security systems everywhere rely on. like it would take 10,000 years to crack an RSA key with conventional computing but quantum could do it in no time or something. that sounds bad. i'm pretty sure the US Department of Defense has already been working on these fault tolerant systems though, for exactly that reason. maybe it already happened.

Link to comment
Share on other sites

18 hours ago, very honest said:

i listened to a podcast a while ago with a guy from DOD who was talking about how quantum computing could nullify encryption "cryptosystems" that military, finance, corporate, and security systems everywhere rely on. like it would take 10,000 years to crack an RSA key with conventional computing but quantum could do it in no time or something. that sounds bad. i'm pretty sure the US Department of Defense has already been working on these fault tolerant systems though, for exactly that reason. maybe it already happened.

The DOD might be in some trouble because they tend to like their old (proven) technology, but some of the newer encryption algorithms are provably just as impossible to crack with quantum computers as they are with normal computers so most systems will probably just adapt.

Link to comment
Share on other sites

Here's the actual source:
https://www.nature.com/articles/s41586-019-1666-5

And a discussion of it:

https://www.nature.com/articles/d41586-019-03213-z

 

Here's a well known quantum computing researcher suggesting that this entire story is, unexpectedly, pretty overblown:

https://www.scottaaronson.com/blog/?p=4372

His tl;dr is that Google claimed to have achieved quantum supremacy by performing a calculation using 53 qubits which would take 10k years on a classical computer, but IBM actually put out a paper saying they could simulate Google's calculation in only 2.5 days, which suggests it's not as impressive as they're trying to make it seem

Link to comment
Share on other sites

Even if they have achieved it, which is still debatable at this point, it's still very far from being useful. Their machine only has 53 qubits, and many of them are used for error correction (basically duplicating parts of the calculation), current technology requires multiple physical qubits for every logical qubit. To do something useful, like break RSA encryption you would need thousands of logical qubits. For other scenarios, like modeling proteins or other physics/chemical uses, you might need 10s of thousands or even millions of qubits. The current qubit technology will probably hit a wall pretty soon, they won't be able to scale it up past a couple hundred qubits at best. There is still a lot of basic research that needs to be done, we're still many decades away from any kind of useful quantum computer.

Link to comment
Share on other sites

Like people talk about AI but there has been like no real AI yet. It’s all fake as shit like talking to an aim chatbot put out to promote a movie in 2009.

And I mean neural networks and deep learning are cool tricks, but it’s really just iterative things that you leave running for months due to hardware limits, because moore’s Law is hitting a wall due to size constraints of our physical world. But being impressed by deep learning is the same thing as being impressed that your computer mined bitcoin for you, it’s just doing what you set it to do. Nbd

however...

quantum computers could be the foundation for some classical AI, like a computer you could really talk to, or an android.

once the line between tech guys tweaking code with one hand, and a real, untethered AI is crossed, then maybe the AI is going to be the thing that figures out new energy solutions, basically unlocks everything that we can’t figure out?

basically we have to create a god.

but there are so many, what seem to me, hard barriers, on the manufacturing level, as far as what our abilities are, that really stand in the way of even building a quantum computer. It’s a close cousin to the old computers that took up a city block. We just can’t even conceive of how these things will be built.

i would enjoy living to see androids, but I just don’t think it gon happen....

/shitpost

Link to comment
Share on other sites

Maybe we can use one to simulate hundreds of models simultaneously to find the optimal model of a sustainable society. Of course, once that's concluded no one will listen to it.

Link to comment
Share on other sites

9 hours ago, sheatheman said:

Like people talk about AI but there has been like no real AI yet. It’s all fake as shit like talking to an aim chatbot put out to promote a movie in 2009.

And I mean neural networks and deep learning are cool tricks, but it’s really just iterative things that you leave running for months due to hardware limits, because moore’s Law is hitting a wall due to size constraints of our physical world. But being impressed by deep learning is the same thing as being impressed that your computer mined bitcoin for you, it’s just doing what you set it to do. Nbd

however...

quantum computers could be the foundation for some classical AI, like a computer you could really talk to, or an android.

once the line between tech guys tweaking code with one hand, and a real, untethered AI is crossed, then maybe the AI is going to be the thing that figures out new energy solutions, basically unlocks everything that we can’t figure out?

basically we have to create a god.

but there are so many, what seem to me, hard barriers, on the manufacturing level, as far as what our abilities are, that really stand in the way of even building a quantum computer. It’s a close cousin to the old computers that took up a city block. We just can’t even conceive of how these things will be built.

i would enjoy living to see androids, but I just don’t think it gon happen....

/shitpost

I'm sorry but this post is really spreading lots of misinformation.  First off, every time a new form of AI is made people say it's not AI anymore and it's just software.  This will likely continue forever.  Chatbots are AI, they're just not strong AI, which is a specific type of AI many people think of as being the whole of AI

 As for neural networks not being impressive, I disagree and there are lots of innovations happening here, I won't go into detail.  Saying that it's not impressive because "it's just what you set your computer to do" is drastically misunderstanding and misrepresenting the field.  There's a reason people who are good at deep learning make 6 figures for introductory level jobs, it's almost universally applicable to countless domains. 

And any deterministic computational process is "just doing what you set it to do".  With machine learning however it's doing things you couldn't personally tell it to do because instead you told it to figure out how to do it itself.  That's pretty impressive

As for quantum computing being the foundation for voice-interaction based human-like AI, there's nothing to suggest this can't be done with classical algorithms and in fact most likely classical methods will be better at these particular tasks than quantum computers, even in the future

8 hours ago, Braintree said:

Maybe we can use one to simulate hundreds of models simultaneously to find the optimal model of a sustainable society. Of course, once that's concluded no one will listen to it.

https://www.scottaaronson.com/blog/

If you take just one piece of information from this blog:


Quantum computers would not solve hard search problems
instantaneously by simply trying all the possible solutions at once

Link to comment
Share on other sites

9 hours ago, Braintree said:

Maybe we can use one to simulate hundreds of models simultaneously to find the optimal model of a sustainable society. Of course, once that's concluded no one will listen to it.

i'm sure they'll try. but they'll probably forget to input some relevant data for the computer to chomp on so its model will be flawed. like how in previous attempts they assumed there was no chaos in nature. also took some time for them to collect more and more data. and they can't ever seem to decide if human are part of nature or meant to control it. 

but i'm just spewing stuff from one of the adam curtis documentaries... ?

 

Link to comment
Share on other sites

we already know the optimal model for sustainable society, it's called stop having babies, stop eating meat, stop travelling using fossil fuels, and plant lots of trees.  this just directly contradicts capitalism so it won't happen.  thus, we need guillotines.  only metaphorically, of course...

Link to comment
Share on other sites

But we already know the answer. It's 42!

Also, it's ironically true that the optimal model for a sustainable society is to have no society. (Stop having babies will be the end of society...)

Link to comment
Share on other sites

Ehm, people have been modelling stock markets and the weather for years. As long as science has existed, id argue. That's what science is all about. Climate science? Models! Cosmology? Models!

As long as you disregard the expected quality of those models, that is. But that's science in a nutshell: you continuously improve models. And since the last century, there has been work on so-called automated theorem provers. The step to automated model generators is relatively small, I'd argue. Again though, a lot of it's value depends on what you'd expect. 

Put differently, you could argue that there's a competition between AI and classical science ( meaning: people theorise and conceptualise first). There will be a point where new theories and concepts will be automatically generated. And ironically, this is already happening to various degrees. Especially in a field like biological medicine where the amount of data is so overwhelmingly big, people need algorithms to understand what's going on. In a way you could argue that in some fields these algorithms are already doing most of the heavy lifting. 

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.