Posted on

DeepSeek: the Chinese AI that crashed the market

The market collapsed following the launch of the R1 version of DeepSeek, an artificial intelligence developed by a Chinese company. What happened?

Over the past few hours, the markets—particularly the NASDAQ (the index of major technology stocks) and the cryptocurrency index—have fallen sharply. Many analysts believe this reaction is due to the launch of the R1 version of DeepSeek, an artificial intelligence system based on language models similar to Chat GPT.

In particular, the speed with which DeepSeek was developed and its extremely low cost caused a stir, especially considering that the model is free and open-source. According to its developers’ statements, the realisation of DeepSeek R1 required only USD 6 million and two months of work.

DeepSeek: a threat to the United States?

What is the leading cause for concern related to this innovation in artificial intelligence, which has contributed to the recent collapse of technology stocks? It is quickly said: DeepSeek seems to work very well, and the costs to develop it are negligible compared to those incurred, for instance, by Google to ‘train’ Gemini ($191 million) or by OpenAI to release Chat GPT 5 (between $1.7 and $2.5 billion). This disparity doubts the robustness of AI-related stocks’ impressive growth.

The most commonly discussed hypothesis—though it should be cautiously approached is that DeepSeek could revolutionize the artificial intelligence market and significantly reduce the demand for specific hardware components. This could potentially lead to a wave of panic selling. Conversely, some argue that this is merely a narrative, a typical ‘catalyst’ used to explain movements that are actually part of normal market fluctuations.

What about the crypto market?

Cryptocurrencies experienced a decline for two primary reasons. First, there is a notable correlation between the stock market and the crypto market: when one market falls, it often pulls the other down as well. Additionally, some analysts believe that macroeconomic factors are at play. For instance, during the Federal Open Market Committee (FOMC) meeting on January 29, interest rates could remain unchanged or even be increased despite the new president, Donald Trump, advocating for a reduction.

The market and price movements

The Nasdaq index experienced a correction of nearly 4% before the market opened, while NVIDIA stock plummeted over 14% in pre-market trading before recovering slightly at the start of the trading session.

In terms of cryptocurrencies, Bitcoin fell below the significant psychological threshold of $100,000, a level considered crucial support by some analysts, but then recovered. Overall ,sentiment regarding the leading cryptocurrency appears steady. Prominent analysts, including Arthur Hayes, continue to predict a price target for Bitcoin between $180,000 and $250,000 during this bull market. Additionally, it’s worth noting that February has historically been a strong month for cryptocurrencies, with Bitcoin typically averaging a performance increase of around 15%.

Buy Bitcoin!

DeepSeek is not a ‘black swan’

Despite the scaremongering and scapegoating regarding the recent drop in prices, many experts believe that DeepSeek should not be considered a ‘black swan.‘ By definition, a black swan refers to unpredictable and disruptive events—such as wars, pandemics, or the unexpected collapse of key sectors or players—that can radically alter markets for a prolonged period. For example, the black swans of the last cycle were the collapse of the Earth-Moon ecosystem and the failure of the centralized exchange FTX.

In the case of DeepSeek, however, we are dealing with an innovation that, while interesting, is likely already reflected in market prices. This is especially true at a time when artificial intelligence is at the forefront of media and financial discussions. When everyone is warning about a potential bubble, it suggests that the information is already widely known and, therefore, largely anticipated.

As several analysts note on social media, a narrative is often constructed to justify periods of panic or sudden sell-offs. Without concrete evidence of a widespread collapse, the current market correction may merely be a technical adjustment within an overall bullish trend. Focusing on fundamentals and long-term prospects is the most prudent strategy in a market known for its volatility, helping investors avoid being swayed by extreme assumptions or temporary ‘noise.’

Posted on

What, indeed, are the risks of artificial intelligence?

What are the risks of artificial intelligence?

What are the risks of artificial intelligence? From privacy to security, from ethical dilemmas to work dislocation

Artificial intelligence and machine learning are incredible technologies with enormous potential and an ocean of use cases we have only explored. Like any invention that has the potential to disrupt the world, the introduction of artificial intelligence into our daily lives also carries risks. This aspect of AI started to emerge in 2022 after the launch of ChatGPT, one of the first AI models to go mainstream.

From job displacement—a phenomenon that describes the future disappearance of certain jobs—to concerns about privacy and security and ethical and social dilemmas that, to date, have only been partially addressed, let us see what the main risks of artificial intelligence are.

The risks of artificial intelligence: machine learning vs deep learning

Before addressing the risks of artificial intelligence in detail, it may be useful to define the concept by specifying the main differences between the various models. First, we can start by defining the goal of artificial intelligence, which is to develop ‘machines’ with machine learning and adaptive capabilities inspired by human learning models.

However, the term artificial intelligence (AI) is often associated with concepts such as deep learning and machine learning (ML), which are considered synonymous even though they actually differ. Machine learning is a sub-area of AI that focuses on developing algorithms that allow computers to learn from data and improve their performance over time without being explicitly programmed for each specific task. ML uses statistics to enable machines to ‘learn’ from data, identifying patterns and making decisions based on past examples. 

Deep Learning, on the other hand, is a more specific subset of machine learning that uses neural networks to learn from data. ChatGPT and Gemini (Google’s AI), are good examples of working deep learning models, albeit still embryonic when considering the potential of this technology.

Finally, before addressing the risks associated with artificial intelligence, we can define the main theories related to it, which are very useful in distinguishing the two most widespread types of AI:

  • Artificial solid intelligence: theory according to which machines will be able to develop self-awareness and thus replicate human intelligence;
  • Weak Artificial Intelligence: theory according to which it is possible to develop machines capable of solving specific problems without being aware of the activities performed.

Artificial intelligence has also found new applications in the cryptocurrency sector in recent years, with numerous innovative projects created to combine the best of these two cutting-edge technologies. On our exchange, you will find a selection of AI cryptos and a Custom Money Box that allows you to buy the four most promising ones in this segment regularly.

Find out now!

The risks of artificial intelligence

Now that we have more precisely defined the concepts that make up AI, we can dive headlong into the central topic of this article, answering the question: What are the risks associated with artificial intelligence? It will be necessary to summarise, although each paragraph in this article should be explored in greater depth in a dedicated article. 

  1. Privacy issues

AI technologies and most social media collect and analyse large amounts of personal data, making privacy an ever-present issue. This issue became even more relevant after the arrest of Telegram CEO Pavel Durov. 

Artificial intelligence is also involved in these concerns. However, privacy management varies greatly depending on the legal jurisdiction. For example, European regulations are much stricter than those in the United States and place greater emphasis on the protection of personal data and the rights of individuals.

  1. Ethical and Moral Dilemmas

The discourse on the ethics of AI systems, especially in decision-making contexts that can have significant consequences, is very complex and convoluted. The main difficulty here lies in translating ethical principles, often subjective and culturally variable, into rules and algorithms that can guide machine behaviour. 

Researchers and developers must give the highest priority to the ethical implications of this technology, not only to prevent potential harm but also to ensure that AI operates in a manner consistent with society’s fundamental values. This requires a constant effort to balance technological innovation and social responsibility.

  1. Safety Risks

In recent years, after artificial intelligence has become mainstream, the security risks associated with its use have risen sharply. Hackers and other malicious actors can exploit AI models to conduct increasingly sophisticated cyber attacks, circumvent existing security measures and exploit system vulnerabilities, putting critical infrastructure and sensitive data at risk.

To mitigate these risks, governments and organisations must develop rigorous best practices for the secure implementation of AI. These concerns not only the adoption of advanced security measures but also the promotion of international cooperation to establish global standards and regulations, which is necessary for many experts in the field. In short, only through a coordinated and proactive approach will it be possible to effectively protect society from security threats arising from the misuse of AI.

  1. Labour displacement

Another risk attributed to artificial intelligence is job displacement, which has the potential to cause significant job losses in several sectors, particularly affecting less skilled workers. Although, according to various research, artificial intelligence and other emerging technologies will be able to create more jobs than they eliminate, the transition will only be difficult. As AI technologies continue to develop and become more efficient, it becomes crucial for the workforce to adapt quickly to these changes.

To remain competitive in a changing landscape, workers need to acquire new skills, with a particular focus on digital and technological skills. This is particularly important for lower-skilled workers, who risk being more vulnerable to dislocation caused by automation. Therefore, retraining and lifelong learning become essential to ensure that the workforce can integrate with, rather than be replaced by, new technologies. Public policies and educational initiatives must support this transition process, providing the necessary tools for workers to adapt and thrive in the AI era.

  1. Disinformation and fake news

Finally, the last risk of artificial intelligence we address in this article concerns fake content generated by this technology, such as deepfakes. Creating this content will make it increasingly easy to deceive even experienced observers, fuelling misinformation and undermining trust in information sources. Combating AI-generated disinformation is essential to preserve the integrity of information in the digital age and to protect the democratic fabric of societies.

A Stanford University study highlighted the urgent dangers of AI in this context, stating that “AI systems are being used in the service of disinformation on the Internet, with the potential to become a threat to democracy and a tool for fascism.” Tools such as deep fake videos and online bots, which manipulate public discourse by simulating consensus and spreading fake news, can harm society in various ways.

These are just some of the risks associated with artificial intelligence and its growing impact on our daily lives, but there are many more to consider. For example, there is a concentration of power in the hands of a few large companies and an increasing dependence on tools based on this technology. Without bordering on science fiction, these problems require attention and concrete solutions. However, it is worth pointing out that AI’s opportunities are sufficiently promising to justify continued investment and development, making the balance between costs and benefits positive overall.

Posted on

Crypto AI: Grayscale launches its ad-hoc fund

Grayscale has just announced its crypto AI fund. Find out what this innovative financial instrument consists of.

Grayscale has just announced its Decentralised AI Fund LLC. This brand-new investment fund will allow those who purchase it to gain exposure to the most important crypto protocols aiming to establish themselves in the artificial intelligence sector. 

What cryptos does this innovative fund consist of? What is Grayscale’s main goal, and what artificial intelligence problems could blockchain solve? Find out in this article.

Discover Crypto AI

Grayscale’s new crypto AI fund

Practically everyone knows Grayscale, mainly because it is the largest native crypto investment fund, the first to launch financial instruments on Ethereum and Bitcoin. For this reason, the news released in the past few hours is essential, given the ability of the team of this cutting-edge financial player to intercept new trends. 

The main problem with artificial intelligence, at least according to Grayscale, concerns the centralisation of the companies that control it. 

Few and far between are those who can offer products that can reach the masses, mainly due to the enormous amount of data they hold. As a solution to this problem, various decentralised AI protocols have emerged, aiming to make their processes even more innovative and intelligent. In particular, blockchain technology makes it possible to distribute the ownership and governance of AI services, thereby increasing transparency.

The cryptocurrencies that make up

For now, the information at our disposal tells us that The Grayscale Decentralised AI Fund will self-rebalance every quarter and will accommodate the following basket of cryptocurrencies:

Buy NEAR, RNDR and FIL

The team has yet to comment on possible future additions, but other cryptos will likely be added over time. Why did Grayscale choose these? Well, because they represent the three main categories of crypto AI around today:

  • Protocols that are building decentralised artificial intelligence services;
  • Projects that seek to solve the main problems encountered by AI platforms;
  • Infrastructure networks and resources required for technology development. For example, decentralised marketplaces for data storage, or those for exchanging GPU computing power and graphics rendering.


To conclude, we can quote the words of Rayhaneh Sharif-Askary, Head of Product & Research at Grayscale, who was mentioned in the press release through which the announcement was made. “The rise of these disruptive technologies has created exciting opportunities for investors, and we believe our crypto AI fund is a great way to invest in this emerging sector. Blockchain-based AI protocols embody the principles of decentralisation, accessibility and transparency and can potentially mitigate the fundamental risks emerging from the proliferation of this technology.”