Elon Musk is about to release a super smart chatbot

Elon Musk hails the upcoming Grok 3 chatbot as an AI model that beats all competitors, and the world will get to see it in the coming weeks.

elon musk xai bloomberg 80456.jpg
Chatbot Grok 3 will be launched by Elon Musk's xAI in the next 1-2 weeks. Photo: Bloomberg

Speaking via video at the World Government Conference held in Dubai on February 13, Elon Musk shared: "Sometimes I think Grok 3 is scary smart."

During the event, he spoke about a variety of topics, including his conflict with OpenAI CEO Sam Altman, his plan to “completely wipe out” some US government agencies, and his predictions about the future of AI.

According to the billionaire, xAI's upcoming model is trained on synthetic data and is capable of reflecting on its mistakes by "digging" through data to achieve logical consistency.

Musk said the company is in the final stages of polishing Grok 3 and could release the chatbot in the next week or two. He doesn’t want to rush the model.

The world's richest man also attacked OpenAI, the company he co-founded with Altman and several other entrepreneurs in 2015.

OpenAI stands for open source, non-profit, but is now working to maximize profits from AI, he said.

He recently offered to buy the nonprofit that controls OpenAI for $97.4 billion, but Altman refused.

The terrifying prospect of ransomware attacks

Security experts are 'extremely concerned about what we will see in 2025' when it comes to ransomware attacks.

171013 how to detect and prevent ransomware attacks 90528.jpg
Ransomware attacks are expected to be much larger in scale by 2025. Photo: brightlineit

Christiaan Beek, senior director of threat analysis at security firm Rapid7, believes that 2024 will be the year of continuous attacks.

The company’s report shows that the number of attacks from ransomware gangs increased sharply last year, with ransoms reaching up to $380 million. The average ransom per case is $200,000.

According to Sebastian Straub, chief solutions architect at security firm N2WS, 2025 promises to be even more dangerous. “I’m extremely concerned about what we’re going to see in 2025… I think we’re going to see some pretty big attacks this year.”

Rapid7's 2024 Ransomware Report found that gangs are sharing technology and source code with each other, indicating interoperability between organizations and, more importantly, increasing sophistication of their cyberattacks.

For each source code, they can create different versions.

Straub notes that using AI to develop attack scenarios leads to more dangerous, more effective attacks. Bad actors also target interconnected systems.

He said sustained global coordination is needed to repel attacks. Without community efforts targeting attackers, change will not happen.

US lawmakers want to ban DeepSeek

Two US lawmakers want to ban government employees from using DeepSeek's AI chatbot on public devices.

deepseek wsj 28193.png
DeepSeek's AI models and applications are receiving great attention globally. Photo: WSJ

Senators Darin LaHood and Josh Gottheimer are preparing to introduce legislation to ban the use of DeepSeek on government devices over national security concerns. DeepSeek is the latest target in the battle for technological supremacy between the United States and China.

If the bill passes, government agencies would have 60 days to come up with standards and guidelines for removing DeepSeek, as well as any other apps from parent company High Flyer, from government devices.

Previously, Australia, Italy and several organizations issued similar bans.

The language in the bill is similar to a law that banned TikTok in the US over concerns that its parent company ByteDance could share US user data with China.

5 new iPhone models to look forward to this year Despite the success of many products, Apple is still most famous for the iPhone. In 2025, there will be 5 new iPhone models launched by Apple, all of which are very much anticipated.