From Supercomputers to the AI Era: The Risk Is Still Exponential

[Clearnet]

So recently Vitalik Buterin
 posted on his X profile a higly technical discussion. Although I'm a technical guy I have atleast a gist of it.

I'm making this short and understandable -

Super Computer/Quantam Computer has the ability to break current Blockchain /Wallet security but still it's too much energy consuming and hard part cause not available for everyone.

But AI could be a different show here, he explains it with a human brain.

There was a time when factoring was a time consuming thing [2^(n/2)]

But over decades our human brain , some smart people found a technique to eliminate the huge number of possibilities at once which is [2^O(n^(1/3))]

The security concern in exactly lying here , what if our AI became so smart and found a completely diff technique and formula so which can take normal computer in 2²⁵⁶ to a couple of hours.

At this point we have a more dangerous threat than Super Computer's. AI is a blessing at the same time it can be a greates threat.


So what's your thoughts on this?


Source -  https://x.com/VitalikButerin/status/2107976296320106851

[JaanusRaim]

What is your solution for that problem ?

[Charles-Tim]

So recently Vitalik Butering

He is called Vitalik Buterin.

That does not means quantum computers are not threats. It can be in the handle of few people or organizations, but that does not mean it can not be used to try to drive bitcoin private key from the public key. Do not be surprised if Google or other organizations use it when it become possible.

AI have been the common threats now, but many developers have started using it as the solution against many of the threats.

[EluguHcman]

what if our AI became so smart and found a completely diff technique and formula so which can take normal computer in 2²⁵⁶ to a couple of hours.

At this point we have a more dangerous threat than Super Computer's. AI is a blessing at the same time it can be a greates threat.

Apparently you are saying the AI is the threat of the Bitcoin Block chain and not literally the technical possessions of the computers.
First note that the threats of the quantum computing has still been a rumour that it contains the hardware with the capacity that can destroy the Bitcoin network security.

I don't know why here you are becoming more influenced of this technological security by the AI.
I know that AI has also been of dangerous tools to break Privacies but that is also limited to certain programs like this Bitcoin because of its combination mechanisms which does not recognize or permit any cryptographic programmer (s) to oversee the Bitcoin and Block chain network.

So I am wondering how the AI too will have that capacity to override the Decentralization and the Bitcoin security.
You should stop over thinking about this in imagining the impossibilities of the monetary system all just because you don't trust technologies.
It is neither not the bank that is safe. It is also good to treat with caution to help yourself out from the uncertainty but bro, life is risky and everything about it does too.

Don't fail to buy Bitcoin when you can because of the threats about the quantum and the AI otherwise, time may come when you regrets why you did not just risk it when you had a good chance to buy.

[X-ray]

But over decades our human brain , some smart people found a technique to eliminate the huge number of possibilities at once which is [2^O(n^(1/3))]

The security concern in exactly lying here , what if our AI became so smart and found a completely diff technique and formula so which can take normal computer in 2²⁵⁶ to a couple of hours.

At this point we have a more dangerous threat than Super Computer's. AI is a blessing at the same time it can be a greates threat.

I don't believe until I'm seeing it myself. Vitalik is just raising concerns about the possibilities but there's no such technique or formula yet that we've known so far.

Maybe closest one is algorithm that is better than shor's to help quantum computer crack faster.

It's all still just a speculation, no concrete proof even though it's indeed should be a concern, but we've been talking about quantum computer being a threat for so long.

[Alvin_talk]

The security concern in exactly lying here , what if our AI became so smart and found a completely diff technique and formula so which can take normal computer in 2²⁵⁶ to a couple of hours.

At this point we have a more dangerous threat than Super Computer's. AI is a blessing at the same time it can be a greates threat.

That is INNOVATION and it is likely not possible because ai is limited in that area. AI as I have always said is a pattern recogniser. Before this will be possible it means there has to be someone who has already developed this technique and the ai model has been trained to master such patterns and techniques. AI can serve as a catalyst that saves time given that the process it is performing is not new (someone or something has performed it in the past, it recognises the pattern and execute almost instantly).

You know what I think? I think marketing strategy of ai models is the problem here. They spread FUD to the end users even in areas they know is beyond ai jurisdiction and capability. Anyone can fall for their hype and fallacies because it's even more convincing now that it has successfully been used to perform some major attack in the crypto space. So I'm not surprised if Vitalik Butering didn't DYOR before spitting out those words. Probably, he made that statement out of FUD. And mind you, it is awkward to compare human brain to any known ai model, they are not the same. One can innovate whereas the other can't.

[Clearnet]

So recently Vitalik Butering

He is called Vitalik Buterin.

Thanks for pointing out this , it was a spelling mistakes of mine.

That does not means quantum computers are not threats. It can be in the handle of few people or organizations, but that does not mean it can not be used to try to drive bitcoin private key from the public key. Do not be surprised if Google or other organizations use it when it become possible.

AI have been the common threats now, but many developers have started using it as the solution against many of the threats.

The focus is not on whether the supercomputer can do this. In just 4 years, AI has developed into something that is now posing a threat. I think now the access to a supercomputer is in a few hands, so we can easily catch the culprit if anything goes wrong. Imagine an algorith can break the whole system if it's in everyone's hands

This could lead a disaster. AI is capable of creating something like this , Vitalik wouldn't bother to write that long article about this thread.

We should look into the future, denying the threat is a mistakes here.

I don't believe until I'm seeing it myself. Vitalik is just raising concerns about the possibilities but there's no such technique or formula yet that we've known so far.

Maybe closest one is algorithm that is better than shor's to help quantum computer crack faster.

People like you did not believe when some smart people used their brains and tried to establish the fact that the Earth was round. Some people even lost their lives to proving this. So Ig in every era, some people like that always exist.

Did you ever think that something like artificial intelligence could do almost all of our work 5 years ago? It was not even a fact. Those who did not fear, did not care, have already been replaced. You are one of them, I can see.

I don't know why here you are becoming more influenced of this technological security by the AI.
I know that AI has also been of dangerous tools to break Privacies but that is also limited to certain programs like this Bitcoin because of its combination mechanisms which does not recognize or permit any cryptographic programmer (s) to oversee the Bitcoin and Block chain network.

Either you did not bother to read my writings, or you did not understand a bit. So tell me there was a time when factoring was so hard for people, right? Then some smart mind finds a way to do it more efficiently, then why are you thinking that the Shor algorithm is the closest thing out there to break the technology?

Smart people don't deny it but accept it. We have yet to uncover many mysteries of our world, so if someone can find a more capable algorithm than Shor's, I wouldn't be surprised. Here comes the help of AI. It is already solving the problems. It can think, it can do work, in future it can do more. Then why are you defying the fact that it can't create a better algorithm in the near future?

That is INNOVATION and it is likely not possible because ai is limited in that area. AI as I have always said is a pattern recogniser. Before this will be possible it means there has to be someone who has already developed this technique and the ai model has been trained to master such patterns and techniques. AI can serve as a catalyst that saves time given that the process it is performing is not new (someone or something has performed it in the past, it recognises the pattern and execute almost instantly).

You know what I think? I think marketing strategy of ai models is the problem here. They spread FUD to the end users even in areas they know is beyond ai jurisdiction and capability. Anyone can fall for their hype and fallacies because it's even more convincing now that it has successfully been used to perform some major attack in the crypto space. So I'm not surprised if Vitalik Butering didn't DYOR before spitting out those words. Probably, he made that statement out of FUD. And mind you, it is awkward to compare human brain to any known ai model, they are not the same. One can innovate whereas the other can't.

Unfortunately, you have very little knowledge about AI. Right now, AI lacks the creativity for which a human intervention is still needed. I am not talking about this happening over time. But in future, this poses a threat. A brilliant mind with a brilliant AI in hand can do 100x faster work.

Do you know AI can self-evolve? You, a human, are there to instruct. So why can't AI develop something that hasn't been figured out yet?

What is your solution for that problem ?

Lol, there is none, cause people are busy arguing that it's not even a threat compared to a supercomputer. They are on their own box could not think the outside box.

[Alvin_talk]

Unfortunately, you have very little knowledge about AI.

Irony 

Do you know AI can self-evolve? You, a human, are there to instruct. So why can't AI develop something that hasn't been figured out yet?

That's what you were taught for the sake of marketing. Give me a real world example of this.

A brilliant mind + ai is effective but ai alone can not innovate any meaningful idea as it stands now. As for the future, we know its pregnant. But going by the procedures involved in developing ai, I don't think it can innovate anything even in the future without a human original idea. It can do whatever we know how to order it to do, with better speed and maybe accuracy. What developers do to build an artificial intelligence (AI) is that they collect and prepare data, then the training/learning stage, then it is fine tuned and evaluated by professionals in diverse fields and I don't think such dependent system will be able to innovate. Unless the foundation will change, else, ai will remain a pattern recognition machine. So before it can become a threat to bitcoin or crypto, there has to be an experienced attacker instructing it how to do it. And there is a solution to such threat; so long as we also have competent developers, we can also use ai to fight security threats from ai.

[Dunamisx]

At this point we have a more dangerous threat than Super Computer's. AI is a blessing at the same time it can be a greates threat.

AI is not a threat to bitcoin network, I may not see much about other cryptocurrencies, but had it been AI stand to be a threat, the meaning of lose hope on bitcoin and seek in for alternative, but the more they keep talking about quantum computers and the threat they are possibly going to have on Bitcoin, the harder it becomes to see this come to a reality because of how secured bitcoin network is.

[Lucius]

A man who will never accept the fact that his shitcoin will forever remain in the shadow of Bitcoin talks about things that don't make sense - because if anything is going to be hacked first, it's his creation. Quantum computers are still a long way from being able to hack Bitcoin, but obviously this story serves those who want to cause panic among people and hopefully convince them that other coins are a better choice.

[Tungbulu]

I’m very much concerned about how the crypto industry is gonna respond if such a remarkable breakthrough actually happened. Let’s even assume (hypothetically of course) that AI actually discovers a potential error or weakness in the current cryptography, the damage would still depend heavily on how fast developers can prepare and move to a much safer alternative. And another challenge would be having to make such a transition before any actual damage as well as to do that without creating new security risks or causing unnecessary panic among several users.

[Multani Boy]

I don't believe until I'm seeing it myself. Vitalik is just raising concerns about the possibilities but there's no such technique or formula yet that we've known so far.

Maybe closest one is algorithm that is better than shor's to help quantum computer crack faster.

It's all still just a speculation, no concrete proof even though it's indeed should be a concern, but we've been talking about quantum computer being a threat for so long.

This does not mean that we need to believe that this technique is possible today. Crucial issue is how quickly we can react should such a technique ever become available. Quantum computing has long been on the horizon, yet an AI discovered technique would be another kind of threat altogether. Monitoring the literature would be a good idea even if we had no evidence that such a technique was possible.

I’m very much concerned about how the crypto industry is gonna respond if such a remarkable breakthrough actually happened. Let’s even assume (hypothetically of course) that AI actually discovers a potential error or weakness in the current cryptography, the damage would still depend heavily on how fast developers can prepare and move to a much safer alternative. And another challenge would be having to make such a transition before any actual damage as well as to do that without creating new security risks or causing unnecessary panic among several users.

One tough issue will be the handling of coins whose owners are no longer active or cannot be reached. Even if devs manage to roll out new scheme fairly quickly, migrating all vulnerable coins might not be easy. Bitcoin will have to plan for such scenarios, and not just focus on how fast we can build new cryptographic solution.

[joniboini]

Even if devs manage to roll out new scheme fairly quickly, migrating all vulnerable coins might not be easy. Bitcoin will have to plan for such scenarios, and not just focus on how fast we can build new cryptographic solution.

So the bigger risk if those coins are untouched are probably market crashing hard. If someone can move or block some of those coins from entering the market in one way or another, I think the bigger issue would be whether the network is really free from a singular control anymore. I can't imagine majority of people agreeing with a new move to tackle "vulnerable coins", especially if it involves some sort of filter that can be changed arbitrarily.

[X-ray]

People like you did not believe when some smart people used their brains and tried to establish the fact that the Earth was round. Some people even lost their lives to proving this. So Ig in every era, some people like that always exist.

Did you ever think that something like artificial intelligence could do almost all of our work 5 years ago? It was not even a fact. Those who did not fear, did not care, have already been replaced. You are one of them, I can see.

My point is I won't believe there is way for classical computer to shorten 2²⁵⁶ until I'm seeing AI generated different technique and formula that that could shorten 2²⁵⁶ to couple of hours in normal computer.

I know and have read the concern about AI speeding up maths, especially the lattice-based cryptography concern floated by Vitalik, but the experts are also already finding ways.

That means im not gonna hole up in a bunker and withdraw all my bitcoin to fiat out of fear.

[notocactus]

So the bigger risk if those coins are untouched are probably market crashing hard. If someone can move or block some of those coins from entering the market in one way or another, I think the bigger issue would be whether the network is really free from a singular control anymore. I can't imagine majority of people agreeing with a new move to tackle "vulnerable coins", especially if it involves some sort of filter that can be changed arbitrarily.

It it comes to a times of "vulnerable bitcoins" because of current and previous wallets, such bitcoins will have to be migrated by owners, not by any automatic migration by Bitcoin developers with any BIPs.

It's my keys, it's my coins, and nobody is allowed to access my coins, to migrate it to anywhere.

I am fully responsible for safety or unsafety of my bitcoins. It's same with the other people and bitcoin owners, if they don't migrate their coins to new more secured chain by new BIP, if they lose their bitcoins, it's their donation to everyone.

The principle works the same with "forked coins" claiming, people are free to claim forked coins or ignore it.

[bitbollo]

P, Ardoino in one of his speech has seen bitcoin as "the monetary unit" for these forms of intelligence, it was last year on 2025 in Lugano Plan B.
These systems can growth fast. "break" blockchain/bitcoin security would be probably a silly experiment for a AGI/AI.
At the end, if they have this "superiority" they could just create a new "bitcoin" or a new sophisticated concept of money.
Worst. They can run a series of scams really hard to detect - endorsed by dozens of bots - that collect bitcoin. The computational costs would be negligible in comparison.
This is the real risk - we are seeing something similar already now, with all these attacks against bitcoin and bitcoiners that are making selfcustody (just to mention one case).
If you consider "Dead Internet theory" (https://en.wikipedia.org/wiki/Dead_Internet_theory) at a certain point you will never see anymore of "humans" on internet.

[Multani Boy]

So the bigger risk if those coins are untouched are probably market crashing hard. If someone can move or block some of those coins from entering the market in one way or another, I think the bigger issue would be whether the network is really free from a singular control anymore. I can't imagine majority of people agreeing with a new move to tackle "vulnerable coins", especially if it involves some sort of filter that can be changed arbitrarily.

Now here is where it can get tricky. We can protect the network against the crypto threat, but saying certain coins should not be spendable is something different.

If you need to trust someone to make a decision about what coins are safe, then you have created a different sort of risk. Any upgrade will need to weigh the security benefits against keeping users in control of their coins.

[Agbamoni]

So what's your thoughts on this?

The chaos, disaster, and future problems in technology, particularly in the area of Bitcoin, are often associated with quantum computing advancements and AI, but they are not what causes it. It is when this technology falls into the wrong hands that leads to chaos and eventually disaster. They are tools that were meant to provide fast solutions and to support humans, but people will rather use them for bad rather than good.

The solution to this problem is for scientists to build a technology that will detect any human around the world planning evil in their mind. A technology that can read evil thoughts from a distance. This will solve major problems in the world.

[5W-KILO]

I honestly like the fact that AI is still creating this much problem for crypto and Blockchain combined, atleast we will all understand how garve danger this really is, I believe that developers are preparing for solutions.

Some projects are building quantum resistance Blockchain and I believe that something will work out for Bitcoin too, super computer can still rest, their time isn't now or else they will create bigger problems.

The world needs to be ready for quantum computers before they are passed to the public, or else even the banking system will be greatly affected.

[Luzin]

At this point we have a more dangerous threat than Super Computer's. AI is a blessing at the same time it can be a greates threat.

So should we turn off AI? AI was created by humans, and Quantum technology is the same. How they develop depends on the capabilities of their creators. Perhaps AI surpasses the abilities of its creators, but that certainly happens in specific areas depending on the ledger that serves as its source. The most important issue is the process of testing and clear limitations to prevent AI from operating outside human control. When that happens is possible, but not in the near future. We cannot reject this technology, if AI can find weaknesses, then AI should also be able to find ways to overcome them. This also applies to Quantum technology. CMIIW