AI Breakthroughs of 2024: From Adorable Robots to Cutting-Edge Technology

1. Casio, together with Vanguard, has released the cutest furry robots, Moflin, which are great stress relievers when you hug and pet them. The fluffy lumps recognize the owner and react to him in a “special” way. And if you try to hug your pet, it will try to snuggle up to you. Each Moflin has a built-in AI that will help you develop a unique character in the toy over time.
2. AI drives Wikipedia into a crisis. Wikipedia is facing a crisis due to the use of artificial intelligence to create articles. Language models generate fictitious information that threatens the reliability of the platform. To solve the problem, the WikiProject AI Cleanup project was created, where volunteers remove fake data.
3. China has developed a police balloon. The robotic bullet is designed to fight crime: it can move at speeds up to 35 km/h and shoot with a special net. The device can not only chase suspects but also swim and work in teams with other similar robots. According to the developers, its body is almost impossible to damage.
4. TikTok is massively laying off employees and replacing them with AI. This is due to the introduction of an automated content moderation system that works much faster than humans and is able to detect and remove up to 80% of prohibited materials on the platform.
5. Youtube will help to distinguish AI videos from natural ones. Now the popular video hosting platform has a mark on videos shot with a real camera, which confirms the absence of neural network interference.
6. The US uses AI to collect money. The US Treasury Department announced that it used neural networks to analyze data and recover $4 billion in “fraud and improper payments” in fiscal year 2024. Thanks to the introduction of AI, the regulator was able to recover almost four times more funds compared to the previous fiscal year.
7. Banks start replacing staff with AI. According to Bloomberg, many European banks said that AI will allow them to reduce costs by performing tasks previously performed by hired employees.
   – Italy’s BPER Banca plans to lay off about 2000 people over the next few years.
   – UBS Group has developed an AI that analyzes databases of more than 300 thousand companies in less than half a minute.
   – Germany’s Deutsche Bank uses AI to scan portfolios of wealthy clients.
8. A group of AI researchers conducted an experiment in which experts tried to “hack” several robots and violate security protocols. As a result, they succeeded and the machine started performing actions contrary to its nature. Experts said that hacking robots is not only possible, but also “alarmingly easy”!
9. Google ordered 7 nuclear reactors for AI. The tech giant has made a historic deal by ordering seven small modular reactors (MMPs) from Kairos Power from California. Google plans to use them to power the growing AI infrastructure. It is also important that the energy source is round-the-clock and low-carbon, which will avoid environmental pollution problems. The location of the reactors and financial terms are still secret. Google will purchase a total of 500 megawatts of electricity from Kairos.
10. An AI painting is sold at auction for the first time. The Ai-Da robot painted a picture depicting Alan Turing, the ancestor of all information technologies. The price of the painting is $180,000. It is on display at the Sotheby’s auction house.
11. Amazon and Microsoft follow Google’s lead by investing in nuclear power. Amazon Web Services (AWS) has announced plans to invest $500 million in small modular reactors (SMRs) to meet growing clean energy needs. The company has entered into agreements with Dominion Energy and Energy Northwest to explore the use of these reactors with the goal of achieving carbon neutrality by 2040. SMRs are compact nuclear reactors capable of producing significant amounts of low-carbon energy. These reactors can be built faster and placed closer to power grids, making them a great solution for AWS data centers. Microsoft has entered into a similar agreement with Constellation Energy to restart a reactor at the Three Mile Island Nuclear Power Plant to supply energy to the company’s data centers.
12. AI can “sabotage” humanity. Anthropic analysts said that one day, AI could “sabotage” humanity – it could pretend to be unable to do something to convince testers that it poses a threat, and it could intentionally allow malicious content to enter as part of a malicious plan. The company studied four different threat vectors of “sabotage” for AI and concluded that “minimal mitigation measures” are sufficient for current models.
13. Apple is not keeping up with the development of AI. The company admitted that the new development of Apple Intelligence is far behind other AI companies. ChatGPT and other neural networks provide much more capabilities than Apple’s upcoming development. Despite the strong advantage of OpenAI in the AI field, Apple did not limit itself to the disadvantages. The company promises to support Apple Intelligence for all major devices, which greatly simplifies and improves the work with the Apple ecosystem.
14. The creators of Blade Runner 2049 filed a lawsuit against Elon Musk for using AI images from the movie to advertise Tesla. It happened at the Cybercab presentation, where Musk did not limit himself to demonstrating the images, but also expressed his admiration for Villeneuve’s film.
15. A leading crypto miner moves into the AI sphere. Northern Data, one of the leading crypto-miners in the EU, backed by Tether, intends to sell its cryptocurrency mining division Peak Mining and switch its focus to artificial intelligence. Earlier it became known that Northern Data is negotiating with consultants to hold an IPO in the United States – the valuation of the combined areas of cloud computing and data processing using AI can range from $10 billion to $16 billion.
16. China Mobile will create a neuro-metaverse. A subsidiary of China Mobile, the world’s largest mobile operator with 1 billion subscribers, plans to create a metaverse based on artificial intelligence.
17. Meta is developing its own AI-based search engine to compete with Google and Microsoft, whose search engines it currently uses to provide users with information about news, stocks, and sports events.
18. OpenAI presented its search engine ChatGPT Search. The new artificial intelligence-based search engine from OpenAI provides fast and accurate answers and links to sources.
19. Musk attracts investments in xAI. Elon Musk is negotiating with investors from the Middle East to finance the artificial intelligence startup xAI, which could result in a $45 billion valuation. Earlier it became known that Musk plans to build a supercomputer called the “Gigafactory of Computing” to support the development of xAI.
20. The first AI cartoon is released. The AiMation film studio has released a full-length cartoon made with the help of neural networks. The cartoon is called Where Robots Grow and is set in the future, where the last inhabitants of the Earth send robots to explore a new planet after a catastrophe.
21. Elon Musk’s chatbot is updated. The xAI company founded by Elon Musk has released an update for its chatbot Grok. The neural network now works directly in the X social network, where users can use it to recognize objects in photos or even explain jokes.
22. SunoAI update. Now the best music neural network can make a song in the spirit of your favorite artist, preserving the mood, musical style, and voice. Available only in the paid version, but the feature is really interesting.
23. Linus Torvalds criticizes AI. The creator of the Linux kernel doesn’t like the hype around artificial intelligence: he says that AI technologies are now “90% marketing and only 10% actual reality.” According to Torvalds, in 5 years, generative algorithms and machine learning technologies will become much more useful and interesting. Then the whole world will be able to understand how AI can be used and what daily tasks it can “speed up”.
24. The new ChatGPT is disappointing. The o1-preview model recently presented by OpenAI showed terrible results, answering only 43% of questions.