BTC $64,964.18 +0.74%
ETH $1,927.88 +1.26%
BNB $587.41 +3.34%
XRP $1.08 +1.03%
SOL $74.62 +1.39%
TRX $0.3280 +0.51%
DOGE $0.0704 -0.01%
ADA $0.1681 +2.36%
BCH $220.49 +4.76%
LINK $8.51 +2.36%
HYPE $53.59 -3.05%
AAVE $99.54 +1.20%
SUI $0.6962 +1.31%
XLM $0.1732 -0.30%
ZEC $476.12 +2.34%
BTC $64,964.18 +0.74%
ETH $1,927.88 +1.26%
BNB $587.41 +3.34%
XRP $1.08 +1.03%
SOL $74.62 +1.39%
TRX $0.3280 +0.51%
DOGE $0.0704 -0.01%
ADA $0.1681 +2.36%
BCH $220.49 +4.76%
LINK $8.51 +2.36%
HYPE $53.59 -3.05%
AAVE $99.54 +1.20%
SUI $0.6962 +1.31%
XLM $0.1732 -0.30%
ZEC $476.12 +2.34%

theo

All
Article
Flash

Claude discovered vulnerabilities in the encryption algorithm, posing a theoretical threat to post-quantum security

Anthropic announced that the Claude Mythos Preview model has made breakthroughs in cryptographic research, discovering improved attack methods against the post-quantum digital signature candidate HAWK. HAWK is a post-quantum signature candidate solicited by NIST to combat quantum computer attacks, which has already passed two rounds of expert review. However, Claude reduced the effective key strength of HAWK-256 from 2^64 to 2^38 in just 60 hours, significantly lowering the theoretical cost of cracking. HAWK has not yet been deployed in practice, and this attack currently does not affect any production systems.Claude also discovered an improved attack against 7-round AES, achieving a speed increase of 200 to 800 times. During the research process, Claude independently completed literature reviews, mathematical reasoning, and experimental validation with limited guidance from cryptographers. The HAWK attack took about 60 hours and had an API cost of approximately $100,000; the AES attack was completed independently by Claude, which generated about 1 billion tokens before proposing core innovative ideas. Anthropic researchers then spent hundreds of hours validating the results. Anthropic stated that AI models can now help identify significant flaws in cryptographic algorithms, and the cryptography community may face bottlenecks similar to those in the software vulnerability field—AI-generated research results far exceed human verification capabilities. Anthropic has collaborated with academic institutions to launch the CryptanalysisBench benchmark and coordinated disclosures with NIST authors and government partners. The research team has achieved preliminary results on algorithms such as LEA and Serpent-128. Anthropic warns that as AI capabilities improve, actual attacks against deployed systems may be discovered in the future, necessitating the establishment of response mechanisms in advance.

Billionaire Dan Loeb refutes the AI bubble theory: The AI investment craze is far from peaking, and massive capital expenditures will yield returns

According to BusinessInsider, billionaire investor and hedge fund founder of Third Point, Dan Loeb, stated in a podcast that current market concerns about the "bubble theory" of artificial intelligence (AI) are greatly exaggerated, and the development stage of the AI industry is completely different from that of the internet bubble period.Loeb pointed out that technology giants, including Alphabet, Microsoft, Amazon, and Meta, have collectively exceeded $700 billion in capital expenditures this year, which may reach $1 trillion next year, with the vast majority allocated for AI infrastructure development. He stated that to believe these capital expenditures will not yield returns is equivalent to thinking that companies are "burning money for no reason," but currently, these companies have strong profitability and ample cash flow, allowing them to support investments with their own balance sheets.Loeb emphasized that this is different from the situation during the internet bubble when "valuations detached from fundamentals," and does not constitute a traditional valuation bubble. He also mentioned that AI companies like Anthropic are experiencing rapid revenue growth and accelerated product applications, indicating that the industry is still in the early stages of expansion.Reports indicate that Anthropic's latest financing valuation is nearing $965 billion, with annualized revenue jumping from $14 billion to $47 billion, further strengthening market confidence in the commercialization potential of AI.However, there are still some investors in the market, including Michael Burry, who express concerns about the overheated valuations of AI, believing that massive investments may struggle to yield corresponding returns. Loeb, on the other hand, stated, "We haven't even scratched the surface of AI development," and believes that we are still in the early stages of long-term growth.

Delphi Digital: Solana's major upgrade Alpenglow is expected to launch in 2026, theoretically reducing confirmation delays by 100 times

Delphi Digital posted on platform X that Solana is preparing for a major upgrade called Alpenglow. This upgrade is a complete overhaul of the consensus mechanism, aiming to achieve sub-second finality by replacing Tower BFT and Proof of History (PoH).Alpenglow introduces two new protocol components: Votor and Rotor. Votor replaces the incremental voting rounds of Tower BFT with a lightweight voting aggregation model. Validators can aggregate votes off-chain before submitting the final confirmation, allowing blocks to achieve finality within 1 to 2 confirmation rounds. This improvement reduces the theoretical finality delay to 100 to 150 milliseconds, shortening it by about 100 times from the original 12.8 seconds.Votor achieves finality through two parallel paths: when a proposed block receives over 80% of the total staked weight in the first round, it triggers fast confirmation and takes effect immediately; if the support rate is between 60% and 80% in the first round, it triggers slow confirmation, requiring a second round of voting to exceed 60% for final confirmation.Rotor reconfigures Solana's block propagation layer. The original Turbine propagation network relied on multi-hop relays with variable delays, while Rotor introduces a staked-weight relay path that prioritizes bandwidth efficiency. Validators with high stakes and reliable bandwidth will become core relay points. Simulation data shows that under typical bandwidth conditions, block propagation can be completed in as fast as 18 milliseconds. This upgrade is expected to be rolled out gradually, with an initial launch timeframe anticipated between early to mid-2026.
app_icon
ChainCatcher Building the Web3 world with innovations.