Foresight published a market analysis article that centered on two issues: whether ETH can still make a new high, and how to think about the possibility that AI could break today’s cryptographic algorithms.
Whether ETH can reach a new high
The author said Ethereum is probably the crypto asset they have written about most, even more than Bitcoin, and added that their earlier articles had already made their personal view on that question fairly clear. Even so, the piece stressed that any outside opinion is only a reference. In the end, an investor’s final decision comes from their own real judgment.
According to the article, people often already know what they actually think, but hesitate to admit it. The hesitation may come from feeling that their true view is out of step, or from worrying that acting on it could mean missing an opportunity. The author’s point was that this kind of internal conflict does not need endless overthinking.
To make that point, the author cited Duan Yongping. The article said the author frequently quotes Duan’s views and sees his methodology as a major source of help, yet still finds it hard to understand Apple, Duan’s largest U.S. stock holding, and has never felt comfortable buying it.
The author wrote that this does not create much conflict, because there are other assets in other markets that feel easier to understand and inspire more confidence. Apple may have brought Duan enormous wealth, the article said, but that does not mean it would do the same for someone else. Even buying along with him would not necessarily produce returns.
How the article views the risk of AI breaking cryptography
The piece said this has become a widely discussed topic recently, and gave two main reasons.
First, OpenAI recently published more than 700 papers at once, with each of them said to claim solutions to mathematical problems that had troubled mathematicians for a long time.
Second, a heavyweight researcher from Ethereum’s core team is concerned that if AI keeps advancing at its current pace, existing cryptographic systems based on elliptic curves could be broken in the future, possibly even before quantum computing reaches large-scale commercial deployment.
The article added that Vitalik also gave a detailed explanation of the second point. Based on Ethereum’s own experience using AI to speed up research and push protocol iteration, he proposed a scenario in which AI could uncover hidden flaws in current cryptographic algorithms that humans have not yet found, and then use them to break those systems.
Putting those points together, the author said the anxiety largely comes from AI’s rapid progress in mathematics.
The article’s breakdown of that concern
The author argued that, when the discussion returns to whether current cryptographic algorithms can actually be broken, the recent leap still has not gone beyond the algorithmic framework itself. What has improved, in this view, is efficiency at the edge of what those algorithms can do.
The piece pointed to two basic elements of current cryptographic systems:
- The private key and public key are different; the private key must remain secret, while the public key is public.
- There is only one mathematical route to compromising a private key: deriving it from the public key that has been exposed.
Under current methods, reversing a public key into a private key would require an enormous amount of computation, which the article said remains beyond what current technology can achieve.
That, according to the piece, is why quantum computing is often described as a threat to private keys. It could sharply reduce the burden of that massive computation and raise efficiency by 100 million times, 1 billion times, or even 10 billion times, making that reversal feasible.
Vitalik’s hypothetical path is different. The article said he suggested that there may be no need for brute-force-style massive computation if existing public keys already contain unknown mathematical patterns. If such patterns exist and can be identified, they might allow a private key to be inferred from a public key. Humans have not found those patterns, the author wrote, but AI might.
The piece said this is the full scope of the author’s understanding of concerns over AI breaking existing cryptographic algorithms.
The mitigation approach described in the article
In the author’s view, the two routes share one condition: both require the full public key to be public.
If the complete public key of an address is not exposed, the article argued, then neither quantum computing nor AI-based methods can derive the private key. The piece described this as the most practical and easiest way available at present to address the concern.
The method itself was presented simply: each time a transfer is made, move all assets to a newly generated address that has never initiated or signed a transaction as a sender. The article said wallets built around this idea have existed for a long time.
It added that this is also why members of Ethereum’s core team recently suggested that users move assets to fresh addresses.
At the same time, the article noted that Vitalik has a different view on using address migration as the answer to the risk. He believes operational mistakes during that process could create problems of their own. As a result, the author said each person should handle the issue according to their own situation.
The article ended by saying that the risk mentioned in the second point, along with Vitalik’s hypothetical scenario, remains only a hypothesis at this stage and has not been proven. Even if it were eventually confirmed, the author wrote, the method described above could still reduce the risk effectively. For that reason, the author said they are not overly worried about it.
The disclaimer at the end of the article stated: "Markets carry risk, and investment requires caution. This article does not constitute investment advice. Users should consider whether any opinion, view, or conclusion in this article fits their specific circumstances. Any investment made on that basis is at their own responsibility."
