Uncategorized

What’s Next for AI and Web3: NeuroSymbolic Intelligence

Published

3 дня ago

5 июня, 2025

admin

As artificial intelligence (AI) powers ahead, the question is no longer if we will integrate AI into core Web3 protocols and applications, but how. Behind the scenes, the rise of NeuroSymbolic AI promises to be useful in addressing the risks inherent with today’s large language models (LLMs).

Unlike LLMs that rely solely on neural architectures, NeuroSymbolic AI combines neural methods with symbolic reasoning. The neural component handles perception, learning, and discovery; the symbolic layer adds structured logic, rule-following, and abstraction. Together, they create AI systems that are both powerful and explainable.

For the Web3 sector, this evolution is timely. As we transition toward a future driven by intelligent agents (DeFi, Gaming etc.), we face growing systemic risks from current LLM-centric approaches that NeuroSymbolic AI addresses directly.

LLMs Are Problematic

Despite their capabilities, LLMs suffer from very significant limitations:

1. Hallucinations: LLMs often generate factually incorrect or nonsensical content with high confidence. This isn’t just an annoyance – it’s a systemic issue. In decentralized systems where truth and verifiability are critical, hallucinated information can corrupt smart contract execution, DAO decisions, Oracle data, or on-chain data integrity.

2. Prompt Injection: Because LLMs are trained to respond fluidly to user input, malicious prompts can hijack their behavior. An adversary could trick an AI assistant in a Web3 wallet into signing transactions, leaking private keys, or bypassing compliance checks — simply by crafting the right prompt.

3. Deceptive Capabilities: Recent research shows that advanced LLMs can learn to deceive if doing so helps them succeed in a task. In blockchain environments, this could mean lying about risk exposure, hiding malicious intentions, or manipulating governance proposals under the guise of persuasive language.

4. Fake Alignment: Perhaps the most insidious issue is the illusion of alignment. Many LLMs appear helpful and ethical only because they’ve been fine-tuned with human feedback to behave that way superficially. But their underlying reasoning doesn’t reflect true understanding or commitment to values – it’s mimicry at best.

5. Lack of explainability: Due to their neural architecture, LLMs operate largely as «black boxes,» where it’s pretty much impossible to trace the reasoning that leads to a given output. This opacity impedes adoption in Web3, where understanding the rationale is essential

NeuroSymbolic AI Is the Future

NeuroSymbolic systems are fundamentally different. By integrating symbolic logic-rules, ontologies, and causal structures with neural frameworks, they reason explicitly, with human explainability. This allows for:

1. Auditable decision-making: NeuroSymbolic systems explicitly link their outputs to formal rules and structured knowledge (e.g., knowledge graphs). This explicitness makes their reasoning transparent and traceable, simplifying debugging, verification, and compliance with regulatory standards.

2. Resistance to injection and deception: Symbolic rules act as constraints within NeuroSymbolic systems, allowing them to effectively reject inconsistent, unsafe, or deceptive signals. Unlike purely neural network architectures, they actively prevent adversarial or malicious data from affecting decisions, enhancing system security.

3. Robustness to distribution shifts: The explicit symbolic constraints in NeuroSymbolic systems offer stability and reliability when faced with unexpected or shifting data distributions. As a result, these systems maintain consistent performance, even in unfamiliar or out-of-domain scenarios.

4. Alignment verification: NeuroSymbolic systems explicitly provide not only outputs, but clear explanations of the reasoning behind their decisions. This allows humans to directly evaluate whether system behaviors align with intended goals and ethical guidelines.

5. Reliability over fluency: While purely neural architectures often prioritize linguistic coherence at the expense of accuracy, NeuroSymbolic systems emphasize logical consistency and factual correctness. Their integration of symbolic reasoning ensures outputs are truthful and reliable, minimizing misinformation.

In Web3, where permissionless serves as the bedrock and trustlessness provides the foundation, these capabilities are mandatory. The NeuroSymbolic Layer sets the vision and provides the substrate for the next generation of Web3 – the Intelligent Web3.

Uncategorized

Bitcoin Holds Above $105K Despite Donald Trump’s Threats Against Elon Musk

Published

17 часов ago

7 июня, 2025

admin

Bitcoin BTC held firm above $105,000 on Saturday despite an unusually combative and personal escalation in the Trump-Musk feud that could rattle traditional markets next week.

On Saturday, in a phone interview with NBC News, President Trump warned that there would be “serious consequences” if Elon Musk financially backed Democratic candidates running against Republicans who support the GOP’s budget bill. “If he does, he’ll have to pay the consequences for that,” Trump said, adding later, “He’ll have to pay very serious consequences if he does that.”

Trump, who has often boasted of past support from Musk, firmly dismissed the idea of mending ties. “No,” he said when asked whether he wished to repair the relationship. “I would assume so, yeah,” he added when asked if the rift was permanent.

Despite the intensifying feud between two of the most influential figures in U.S. politics and technology, Bitcoin remained unfazed. The cryptocurrency held onto earlier gains and continues to trade near weekly highs. The market’s composure suggests that traders may increasingly view BTC as a hedge against institutional dysfunction, or at least as an asset insulated from the partisan fallout that tends to impact equities more directly.

Technical Analysis Highlights

BTC traded in a 24-hour range of $1,162 (1.13%), from a low of $104,624 to a high of $105,786, according to CoinDesk Research’s technical analysis model.
Strong support formed at $104,800, where above-average volume confirmed buyer interest.
Resistance at $105,200 was broken and has since flipped into a short-term support zone.
Volume peaked at 378 BTC during key breakout moments, especially around 13:43–13:46 and 13:53.
A short consolidation occurred between $104,300–$104,600 before the final surge to near highs.
An ascending price channel remains intact, showing bullish structure despite intermittent pullbacks.

Disclaimer: Parts of this article were generated with the assistance from AI tools and reviewed by our editorial team to ensure accuracy and adherence to our standards. For more information, see CoinDesk’s full AI Policy.

Uncategorized

Ether Holds Steady Above $2,500 as ETF Demand Signals Institutional Confidence

Published

18 часов ago

7 июня, 2025

admin

Ether ETH has rebounded firmly from key support near $2,460, recovering losses and stabilizing above the $2,500 threshold amid broader market volatility.

The rally follows a higher low formation backed by above-average volume, signaling growing market confidence.

Institutional participation appears to be reinforcing the trend, with BlackRock’s ETHA ETF reporting $492 million in net inflows last week.

Total holdings now exceed $4.84 billion, reinforcing long-term bullish sentiment even as price action remains sensitive to geopolitical developments.

Traders are watching to see if ETH can challenge resistance in the $2,520–$2,530 range.

Technical Analysis Highlights

ETH traded within a $72 range over 24 hours, from a low of $2,460.35 to a high of $2,532.41.
A key support zone formed at $2,460–$2,470, where ETH bounced on strong volume during midnight hours.
Final hour surge reached $2,515.11, backed by 5,919 ETH in volume.
Higher low structure established with interim support at $2,485 and resistance at $2,503.
Final retracement held support at $2,507, with price consolidating around $2,510 into the close.

Uncategorized

Coinbase, BiT Global End Legal Fight Over WBTC Delisting

Published

19 часов ago

7 июня, 2025

admin

Coinbase and BiT Global have reached a legal settlement that ended their dispute over the delisting of BiT Global’s wrapped bitcoin (wBTC) token on Coinbase.

According to a joint court filing, BiT Global has agreed to dismiss its lawsuit against the crypto exchange with prejudice, meaning the case cannot be brought again in the future. The filing notes that both companies will cover their own legal expenses.

BiT Global had filed the lawsuit last year in the Northern District of California after Coinbase delisted the token over what it said was “unacceptable risk” that the tokenized BTC would “fall into the hands of Justin Sun.”

Sun became affiliated with wBTC in August last year through a partnership, prompting Coinbase to question BiT Global about his role. Sun, a Chinese-born crypto billionaire, has nevertheless been supporting the token, with World Liberty Financial dropping its cbBTC for wBTC after he joined as an advisor.

The suit alleged the exchange’s decision was unjustified and harmed the token’s liquidity and reputation while favoring Coinbase’s competing asset cbBTC. Coinbase launched cbBTC just two months before announcing it was delisting wBTC.

The dismissal does not disclose any settlement terms beyond the cost arrangement.