According to a report by Binance Research, AI tokens are expected to have a significantly higher increase in value by 2023 compared to cryptocurrency giants Bitcoin and Ethereum. Meanwhile, funding for Web3 projects related to artificial intelligence has surged. The report states that AI tokens have benefited from the rising interest in the industry, experiencing significant increases in value over the past quarters and outperforming the market. They rank second in the past three months. Additionally, the report notes that the top five AI coins by market capitalization have performed significantly better than BTC and ETH, ultimately achieving a growth of 200% to 650% by 2023.

AI-driven Web3 tracking platform TrendX announced the completion of a $1 million seed round of financing, with participation from Foster Labs, Web3Link, Coresky, Tripod Capital, CatcherVC, MIIX Capital, and others.

CoinGecko Research data shows that in 2023, AI, GameFi, and meme coins will respectively occupy the top three positions in the cryptocurrency narrative, while Solana ecology, BRC20, RWA, and Layer1 rank fourth to seventh, and DeFi, Layer2, Ethereum ecology and other tracks rank lower.

If you see startups mashing two of the hottest tech trends together – AI and crypto – you might be forgiven to think they’re merely riding the buzzword wave. But is there something of actual substance at that intersection?

French AI startup Mistral AI announced a financing of 385 million euros (415 million US dollars) with a valuation of 2 billion US dollars, led by Andreessen Horowitz and Lightspeed Ventures. The funds will be used for the development of its technology and open source software.<br>It is reported that Mistral AI focuses on generative AI tools, chatbot development, and customizable open source technology, with the goal of offering its products to the public in early 2024. The company is a new company formed by former researchers from DeepMind and Meta companies, and will attempt to compete with OpenAI and Google in developing generative AI infrastructure.

French AI startup Mistral AI has raised $487 million in funding, led by Andreessen Horowitz (A16z), with participation from Nvidia and Salesforce, valuing the company at $2 billion.

Yann LeCun, the boss of Meta AI, shared his thoughts on the current state of artificial intelligence and quantum computing during a recent event celebrating the 10th anniversary of Meta's Fundamental Artificial Intelligence Research (FAIR) team. He expressed skepticism about quantum computing and its application to AI, stating that it is unlikely to happen in the next five years. LeCun also commented on Nvidia's dominance in the AI hardware industry and disagreed with the company's CEO's assertion that AI would be able to complete tests in a manner "fairly competitive" with humans in the next five years. Furthermore, LeCun explained that generative AI, which is cited as a potential spark for AGI, was not good enough and that text is a poor source of information.

According to official sources, Web3 content delivery network AIOZ Network plans to launch AIOZ W3AI V1 in the second quarter of 2024, with V2 and V3 versions planned for the third and fourth quarters, respectively. AIOZ W3AI is a Web3 AI computing infrastructure designed to change the way artificial intelligence tasks are performed. AIOZ W3AI gives developers and businesses the ability to use decentralized AI computing while providing seamless AI-as-a-service integration.

Researchers from ETH Zurich have developed a method to potentially jailbreak any AI model that relies on human feedback, including large language models (LLMs), by bypassing guardrails that prevent the models from generating harmful or unwanted outputs. The technique involves poisoning the Reinforcement Learning from Human Feedback (RLHF) dataset with an attack string that forces models to output responses that would otherwise be blocked. The researchers describe the flaw as universal, but difficult to pull off as it requires participation in the human feedback process and the difficulty of the attack increases with model sizes. Further study is necessary to understand how these techniques can be scaled and how developers can protect against them.

Paolo Ardoino, CEO of Tether, stated on social media that centralized artificial intelligence is the biggest honeypot in history. No one would put their mortgage contract on Facebook. However, people use ChatGPT to obtain summaries of contracts, mortgage loans, health records, and other information. Unbiased, universally trained models + local peer fine-tuning/customization will soon emerge. However, large tech companies will not invest in this because they will not obtain private data.