Qubic’s AIGarth AI Fails Basic Math as Monero 51% Attack Claims Spur Security Concerns

Qubic AIGarth AI—trained during a 51% Monero attack attempt—fails basic math, raising security concerns and sending Qubic’s token lower.

Qubic — a proof-of-work blockchain that splits miner resources between Monero mining and AI training — has publicly launched its latest model, AIGarth. Rather than a technical milestone, the rollout has drawn attention for the model’s surprising inability to solve simple arithmetic and for its connection to Qubic’s attempted 51% attack on privacy coin Monero.

On X, the model often replies with a single period and has produced incorrect results such as ‘1+1 = -114’ and ‘1+2 = 28’, prompting public ridicule. Qubic founder Sergey Ivancheglo defended AIGarth’s behavior, saying the model attempts to deduce answers instead of recalling memorized facts — an approach he argues is closer to genuine intelligence.

Qubic describes its training process as Useful Proof-of-Work: miners use compute to generate artificial neural networks (ANNs); a ‘Teacher’ ANN evaluates and improves those networks, then trains subsequent Teachers in a recursive chain intended to move toward artificial general intelligence (AGI).

The AI release follows Qubic’s widely reported claim that it mounted a 51% attack on Monero. Independent analysis said the effort ultimately fell short, but the effort continues to alarm the Monero community. The episode also coincided with a roughly 7.7% drop in Qubic’s token over the past week to an approximate $266 million market cap.

Monero developers are discussing countermeasures, including introducing masternodes or ChainLocks-style protections, or more radical changes such as moving to proof-of-stake. If mining resources remain concentrated or dual-used for offensive operations, the network faces increased security and centralization risks, and token volatility may follow.

Ivancheglo warned critics not to underestimate the project, saying: ‘Real AI isn’t one which passes the Turing test, but one which fails it intentionally.’

Why this matters: the episode combines two core risks for crypto communities — experimental AI development tied to mining and a real-world attack vector against a privacy coin. Users, node operators, and investors should monitor network upgrades and token movement as defenses and market reactions evolve.

Source: Decrypt. Read the original coverage for full details.

Total
0
Shares
Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts